N O V E M B E R 3 , 2 0 2 5 Positive clinical updates on Caribou’s off-the-shelf CAR-T cell therapy programs: • Vispa-cel for r/r LBCL (vispacabtagene regedleucel; CB-010) • CB-011 for r/r MM 2 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. Important information Forward-looking statements This presentation contains forward-looking statements (FLS) within the meaning of the Private Securities Litigation Reform Act of 1995. These forward-looking statements are subject to a number of known and unknown risks, assumptions, uncertainties, and other factors that may cause the actual results, levels of activity, performance, or achievements of Caribou Biosciences, Inc. (the “Company,” “Caribou,” “we,” or “its”) to be materially different from those expressed or implied by any forward-looking statements. In some cases, you can identify forward-looking statements by terms such as “may,” “will,” “should,” “expect,” “likely,” “plan,” “anticipate,” “could,” “intend,” “target,” “project,” “contemplate,” “believe,” “estimate,” “predict,” “potential,” “contribute to,” or “continue” or the negative of these terms or other similar expressions, although not all forward-looking statements contain these words. All statements, other than statements of historical facts contained in this presentation, are forward-looking statements, including but not limited to any statements regarding the initiation, timing, progress, strategy, plans, objectives, expectations (including as to the results) with respect to the Company’s CAR-T cell therapy product candidate clinical trials, including its expectations regarding the timing of initiating dose expansion by the end of 2025 and reporting dose expansion data, along with longer follow-up data on dose escalation, in 2026 from its ongoing CaMMouflage Phase 1 clinical trial for CB-011 in patients with relapsed or refractory multiple myeloma; the expected design, protocol, and timing of initiation of the pivotal phase 3 clinical trial for vispa-cel in 2L LBCL CD19-naïve patients; its ability to successfully develop its CAR-T cell therapy product candidates and to obtain and maintain regulatory approval for these product candidates; the projected manufacturing costs for its CAR-T cell therapy product candidates; the potential commercial opportunities for its CAR-T cell therapy product candidates; the likelihood of its clinical trials demonstrating safety and efficacy of its CAR-T cell therapy product candidates; the beneficial characteristics, safety, efficacy, therapeutic effects, and potential advantages of its CAR-T cell therapy product candidates; the expected timing or likelihood of regulatory filings and approval for its CAR-T cell therapy product candidates; its expected cash, cash equivalents, and marketable securities as of September 30, 2025; and the sufficiency and anticipated use of its existing capital resources to fund its future operating expenses and capital expenditure requirements and needs for additional financing. As a result of many factors, including but not limited to, risks related to its limited operating history, history of net operating losses, financial position, and its need for and ability to raise substantial additional capital to fund its operations and CAR-T cell therapy product candidate development including the ability to fully fund its pivotal phase 3 trial for vispa-cel, and the potential dilution to its stockholders resulting therefrom; risks associated with the initiation, cost, timing, progress, and results of current and future clinical trials, including risks associated with the manufacturing of its product candidates; the risk that initial, preliminary, or interim clinical trial data will not ultimately be predictive of the safety and efficacy of its CAR-T cell therapy product candidates or that clinical outcomes may differ as patient enrollment continues and as more clinical data becomes available or different conclusions or considerations are reached once additional data have been received and fully evaluated; risks related to its ability to obtain and maintain regulatory approval for its product candidates; risks of it not being ult imately able to commercialize its product candidates; risks that its product candidates, if approved, may not gain market acceptance due to negative public opinion and increased regulatory scrutiny of cell therapies involving genome editing; risks related to its ability to meet future regulatory standards with respect to its products; risks related to the substantial uncertainty regarding the current U.S. Administration’s initiatives and how these might impact the U.S. Food and Drug Administration (the “FDA”) and other government agencies in their implementation of laws, regulations, policies, and guidance; risks related to its ability to establish and/or maintain intellectual property rights covering its product candidates and genome-editing technology; risks of third parties asserting that its product candidates infringe their patents; risks related to developments of its competitors and its industry; risks related to its reliance on third parties to conduct its clinical trials and manufacture its product candidates; risks caused by public health crises or geopolitical events on its business and operations; risks related to the volatili ty of its stock price and its potential failure to meet the continuing listing requirements of Nasdaq; and other risks described in greater detail in its filings with the Securities and Exchange Commission (the “SEC”), including the section titled “Risk Factors” of its Annual Report on Form 10-K for the year ended December 31, 2024, and other filings the Company makes with the SEC, the events and circumstances reflected in its forward-looking statements may not be achieved or may not occur, and actual results could differ materially from those described in or implied by the forward-looking statements contained in this presentation. As a result of these risks, you should not place undue reliance on these forward-looking statements. The forward-looking statements in this presentation are made only as of the date hereof. Except to the extent required by law, the Company assumes no obligation and does not intend to update any of these forward-looking statements after the date of this presentation or to conform these statements to actual or revised expectations. Inherent limitations of comparisons with other immunotherapies Caution should be exercised when interpreting results from separate trials involving other immunotherapies. The clinical trial results of other immunotherapies presented or referenced in these slides have been derived from publicly available reports of clinical trials not conducted by the Company, and the Company has not performed any head-to-head trials comparing any of these other immunotherapies with vispa-cel or CB-011. As such, the results of these other clinical trials may not be comparable to clinical results for vispa-cel or CB-011 and may not accurately reflect the true relative efficacy and safety advantages of vispa-cel or CB-011 in comparison to the other immunotherapies presented. The designs of these other trials vary in material ways from the design of the clinical trial for vispa-cel or CB-011, including with respect to patient populations, follow-up times, the clinical trial phase, and subject characteristics. Most of the other trials presented or referenced in these slides have greater patient populations and patient cohorts and longer follow-up times. Accordingly, it is possible that when vispa-cel or CB-011 is evaluated in equally large patient populations over an equally long time period, their safety and efficacy benefits relative to other immunotherapies may be diminished or eliminated. As a result, cross-trial comparisons may have no interpretive value on vispa-cel or CB-011’s existing or future results. For further information and to understand these material differences, you should read the reports for the other immunotherapies’ clinical trials and the sources included in this presentation. This presentation discusses product candidates that have not yet been approved for marketing by the FDA. No representation is made as to the safety or effectiveness of these product candidates for the therapeutic uses for which they are being evaluated. From time to time, the Company may release additional data from its ANTLER phase 1 clinical trial and its CaMMouflage phase 1 clinical trial. The Company makes no representations regarding such additional clinical data or the timing of its release, or whether any such data will support or contradict the findings of any clinical data reported earlier. This presentation shall not constitute an offer to sell or the solicitation of an offer to buy any securities. 3 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. 1 Initial safety and efficacy data on the confirmatory cohort (20 patients) with partial HLA matching, with a minimum of six months of follow up for the majority of patients, as well as an update on the larger, maturing dataset presented previously. Two exciting clinical readouts to share with you today Opening remarks Rachel Haurwitz, PhD, President and CEO, Caribou Biosciences ANTLER Phase 1 trial data Mehdi Hamadani, MD, Medical College of Wisconsin Pivotal trial study design Tina Albertson, MD, PhD, CMO, Caribou Biosciences CaMMouflage Phase 1 dose escalation data Adriana Rossi, MD, Mount Sinai Discussion with clinicians Tina Albertson, MD, PhD, CMO, Caribou Biosciences Joseph McGuirk, DO, University of Kansas Cancer Center Adriana Rossi, MD, Mount Sinai Concluding remarks Q&A 4 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. Caribou has the blueprint for allogeneic CAR-T cell therapies Armoring for functional persistence chRDNA for precision genome editing We leveraged our large clinical data set to identify key attributes Partial HLA matching Donor age 140 patients dosed Vispa-cel (CB-010) efficacy and durability on par with autologous CAR-Ts; safety unlocks outpatient use chRDNA: CRISPR hybrid RNA-DNA; HLA: human leukocyte antigen Vispa-cel CB-011 Anti-CD19 targeting Checkpoint disruption Anti-BCMA targeting Immune cloaking CB-011 drives deep, durable responses with best-in-class allo CAR-T potential for r/r multiple myeloma >

5 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. Vispa-cel: delivering on the allogeneic CAR-T cell therapy promise 1Based on previously reported data from approved autologous CAR-T therapies; Caribou has not performed any comparative analysis directly with such therapies (see Important Information) 2Perales, M-A, et al. Poster 549, 2025 Tandem Meetings 3Mikhael, J. et al. JCO Oncology Practice 2022 18:12, 800-807 4Data on file 2L: second-line; CDMO: contract development and manufacturing organization; LBCL: large B cell lymphoma Autologous CAR-Ts1 1 dose per manufacturing batch Weeks to months for treatment3 Vispa-cel ~75% of 2L LBCL patients do not receive auto CAR-Ts2 A cc e ss S p e e d S ca le Sufficient yield for 200-300 doses per manufacturing batch Eligibility to treatment on the same day4 Many more patients could be served with off-the-shelf CAR-T cells M fg Multiple manufacturing plants One 500 ft2 suite at a CDMO Potential for 96% lower COGS than current autologous CAR-Ts 6 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. Cohort criteria and breakdown of patient numbers 2L: second-line; 3L+: third line or later; HLA: human leukocyte antigen; LBCL: large B cell lymphoma; yo: years old Confirmatory cohort N=22 • CD19n • 2L LBCL • 80M dose level • 4+ HLA matching 7 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. Cohort criteria and breakdown of patient numbers 2L: second-line; 3L+: third line or later; HLA: human leukocyte antigen; LBCL: large B cell lymphoma; yo: years old Confirmatory cohort N=22 • CD19n • 2L LBCL • 80M dose level • 4+ HLA matching Optimized cohort N=35 • CD19n • LBCL o 2L (N=32) o 3L+ (N=3) • 40M, 80M, 120M dose levels • 2+ HLA matching • Young donor, <30 yo 8 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. Cohort criteria and breakdown of patient numbers 2L: second-line; 3L+: third line or later; HLA: human leukocyte antigen; LBCL: large B cell lymphoma; yo: years old Confirmatory cohort N=22 • CD19n • 2L LBCL • 80M dose level • 4+ HLA matching Optimized cohort N=35 • CD19n • LBCL o 2L (N=32) o 3L+ (N=3) • 40M, 80M, 120M dose levels • 2+ HLA matching • Young donor, <30 yo 20 pts from confirmatory cohort w/ 4+ HLA and young donor 2 pts old donor

9 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. Cohort criteria and breakdown of patient numbers 2L: second-line; 3L+: third line or later; HLA: human leukocyte antigen; LBCL: large B cell lymphoma; yo: years old Confirmatory cohort N=22 • CD19n • 2L LBCL • 80M dose level • 4+ HLA matching Optimized cohort N=35 • CD19n • LBCL o 2L (N=32) o 3L+ (N=3) • 40M, 80M, 120M dose levels • 2+ HLA matching • Young donor, <30 yo 20 pts from confirmatory cohort w/ 4+ HLA and young donor 15 pts from dose escalation/expansion w/ 2+ HLA and young donor 2 pts old donor 10 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. Vispa-cel drives deep, durable responses, demonstrating best-in-class allogeneic CAR-T cell therapy potential for r/r LBCL 86% ORR 53% 12-month PFS 63% CR rate No GvHD or Gr 3+ ICANS, <5% grade 3+ CRS, and manageable rates of infections and prolonged cytopenias4 Efficacy and durability on par with autologous CAR-T cells1 Potential best-in-class allogeneic CAR-T cell therapy for safety, efficacy, and durability with optimized vispa-cel3 Pivotal trial in 2L LBCL Expected trial design2: randomized, controlled trial in CD19-naïve, auto CAR-T- and transplant-ineligible patients; control arm to be treated with investigator choice of standard of care immunochemotherapy regimens 1Based on previously reported data from approved autologous CAR-T cell therapies; Caribou has not performed any comparative analysis directly with autologous CAR-T cell therapies (see Important Information) 2Pivotal study approach based on interactions with the FDA to date; the Company intends to further refine the pivotal trial design through continued engagement with the FDA prior to initiation of pivotal trial 3N=35; CD19 naïve, LBCL patients treated with 40M, 80M, or 120M vispa-cel CAR-T cells optimized for multiple factors, including 2+HLA matched and young donor 4Prolonged cytopenias are defined as Grade 3 or 4 neutropenia, thrombocytopenia, or anemia ongoing at day 28 (+/- 5 days) post CAR-T infusion, based on laboratory data, distinct from investigator-reported clinical adverse events. 2L: second-line; CR: complete response; CRS: cytokine release syndrome; GvHD: graft versus host disease; ICANS: immune effector cell-associated neurotoxicity syndrome; ORR: overall response rate; PFS: progression-free survival; r/r: relapsed or refractory Efficacy data cutoff 29Sept2025; safety data cutoff 02Sept2025 11 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. Presenting ANTLER Phase 1 clinical trial data Disclosures Research support/funding: ADC Therapeutics, Spectrum Pharmaceuticals, Astellas Pharma Consultancy: Autolus, Forte Biosciences, Byondis, Kite, Daiichi Sankyo, BMS, Caribou Speaker’s bureau: AstraZeneca, ADC Therapeutics, BeiGene, Kite, Sobi DMC: Myeloid Therapeutics (2023), CRISPR (2024) Professor of medicine and section chief of hematologic malignancies and investigator for the ANTLER trial Medical College of Wisconsin Mehdi Hamadani, MD 12 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. 84 patients dosed with vispa-cel in ANTLER Phase 1 trial -9 to -3 DAYS DAY 0 28 DAYS 3 MONTHS 6 MONTHS 9 MONTHS 12 MONTHS Lymphodepletion Safety and tolerability Response assessment Cyclophosphamide (60 mg/kg/d for 2 days) followed by fludarabine (25 mg/m2/d for 5 days)3 SINGLE DOSE of vispa-cel NCT04637763 1B-NHL subtypes include: DLBCL (diffuse large B cell lymphoma ), HGBL (high-grade B cell lymphoma), tFL (transformed DLBCL from follicular lymphoma), PMBCL (primary mediastinal large B cell lymphoma), FL (follicular lymphoma, with POD24 (high risk)), MCL (mantle cell lymphoma), MZL (marginal zone lymphoma) 2LBCL subtypes include: DLBCL NOS (not otherwise specified), HGBL, transformed DLBCL from FL or MZL, and PMBCL 3Clin Cancer Res. 2011 July 1; 17(13): 4550–4557. doi:10.1158/1078-0432.CCR-11-0116 2L: second-line; B-NHL: B cell non-Hodgkin lymphoma: LBCL: large B cell lymphoma; r/r relapsed or refractory ANTLER trial design for all cohorts Eligibility • Dose escalation: aggressive r/r B-NHL1 with ≥2 prior lines of chemoimmunotherapy or primary refractory • Dose expansion: second-line LBCL2 Exclusion • Prior CD19-targeted therapy for CD19 naïve cohorts Part of trial Patient population N CD19 naïve Dose escalation r/r B-NHL 16 Yes Dose expansion 2L LBCL 41 Yes Confirmatory cohort 2L LBCL 4+ HLA match 22 Yes CD19 relapsed LBCL 5 No

13 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. Baseline characteristics for patients enrolled in ANTLER Patient and disease characteristics All patients N=84 Confirmatory cohort1 N=22 Optimized profile2 N=35 Age, years, median (range) 66 (20-86) 61 (20-83) 63 (20-86) Male, n (%) 64 (76) 16 (73) 25 (71) ECOG, n (%) 0 40 (48) 13 (59) 19 (54) 1 44 (52) 9 (41) 16 (46) NHL subtype, n (%) DLBCL, NOS 48 (57) 14 (64) 21 (60) HGBL 13 (15) 4 (18) 5 (14) tFL 14 (17) 4 (18) 7 (20) tMZL 1 (1) - 1 (3) PMBCL 2 (2) - 1 (3) MCL 3 (4) - - FL 2 (2) - - MZL 1 (1) - - Prior lines of therapy, n (%) 1 67 (80) 22 (100) 32 (91) 2+ 17 (20) 0 (0) 3 (9) Age-adjusted IPI (%) 0-1 33 (39) 13 (59) 18 (51) 2 32 (38) 5 (23) 10 (29) 3 19 (23) 4 (18) 7 (20) Baseline LDH status (%) > ULN 33 (39) 10 (45) 16 (46) > 2x ULN 13 (15) 1 (5) 2 (6) Bulky disease3 17 (20) 2 (9) 4 (11) 12L LBCL 4+ HLA matched, dosed with 80M vispa-cel CAR-T cells 22L (N=32) and 3L+ (N=3) LBCL patients treated with 40M, 80M, and 120M vispa-cel CAR-T cells optimized for multiple factors, including 2+ HLA matched and young donor 3Bulky disease defined by maximum baseline lesion diameter ≥7.5 cm Data cutoff 02Sept2025 2L: second-line; DLBCL, NOS: diffuse large B cell lymphoma, not otherwise specified; ECOG: Eastern Cooperative Oncology Group; FL: follicular lymphoma; HGBL: high- grade B cell lymphoma; HLA: human leukocyte antigen; IPI: International Prognostic Index; LDH: lactate dehydrogenase; MCL: mantle cell lymphoma; MZL: marginal zone lymphoma; NHL: non-Hodgkin lymphoma; PMBCL: primary mediastinal B cell lymphoma; tFL: transformed follicular lymphoma; tMZL: transformed marginal zone lymphoma; ULN: upper limit of normal 14 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. Optimized vispa-cel profile results in efficacy and durability on par with auto CAR-Ts Optimized product includes young donor T cells and 2+ HLA matching Vispa-cel Axi-cel ZUMA-73,4 N=180 Liso-cel TRANSFORM5,6 N=92 Confirmatory cohort 1 N=22 Optimized profile2 N=35 ORR 82% 86% 83% 86% CR rate 64% 63% 65% 66% Median PFS7 (95% CI) NR (2.0, NE) NR (2.8, NE) 14.9 mo (7.2, NE) 14.8 mo (6.6, NR) 12-month PFS (95% CI) 51% (28, 70) 53% (34, 69) 54% (45.8, 60.7) 52% (35.8, 66.4) Median DoR8 (95% CI) NR (1.7, NE) NR (2.1, NE) 26.9 mo (13.6, NE) 12.6 (5.7, NR) FOR ILLUSTRATIVE PURPOSES ONLY. No head-to-head trials between these products have been conducted. Caution is advised when comparing results of different clinical studies as there are differences in patient populations, follow-up times, clinical trial phases, subject characteristics, trial design, and other factors. See Important Information. 12L LBCL 4+ HLA matched, dosed with 80M vispa-cel CAR-T cells 22L (N=32) and 3L+ (N=3) LBCL patients treated with 40M, 80M, or 120M vispa-cel CAR-T cells optimized for multiple factors, including 2+ HLA matched and young donor 3Yescarta prescribing information 4Yescarta FDA Statistical Review 1Apr2022 5Breyanzi prescribing information 6Breyanzi FDA Statistical Review 24Jun2022 7Median follow up 6.0 mo for confirmatory; 11.8 mo for optimized; 22.1 mo for ZUMA-7; 6.2 mo for TRANSFORM 8Median follow up 5.1 mo for confirmatory; 7.9 mo for optimized; NR for ZUMA-7; 4.3 mo for TRANSFORM Note: For axi-cel and liso-cel, efficacy data is from respective study IRC (Independent Review Committee) CR: complete response; DoR: duration of response; HLA: human leukocyte antigen; NE: not evaluable; NR: not reached; mo: month; ORR: overall response rate; PFS: progression-free survival Efficacy data cutoff 29Sept2025 15 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. Optimized vispa-cel profile results in efficacy and durability on par with auto CAR-Ts Optimized product includes young donor T cells and 2+ HLA matching Confirmatory Optimized profile 1.0 0.8 0.6 0.4 0.2 0.0 +censored Number at risk 22 12 7 4 3 2 0 35 24 16 13 11 8 6 6 6 1 1 1 1 1 0 S u rv iv a l p ro b a b il it y 0 12 24 36 48 60 72 84 96 108 120 132 144 156 168 Progression-free survival Post vispa-cel infusion (weeks) Liso-cel: 14.8 mo median PFS Axi-cel: 14.9 mo median PFS FOR ILLUSTRATIVE PURPOSES ONLY. No head-to-head trials between these products have been conducted. Caution is advised when comparing results of different clinical studies as there are differences in patient populations, follow-up times, clinical trial phases, subject characteristics, trial design, and other factors. See Important Information. Vispa-cel Axi-cel ZUMA-73,4 N=180 Liso-cel TRANSFORM5,6 N=92 Confirmatory cohort 1 N=22 Optimized profile2 N=35 ORR 82% 86% 83% 86% CR rate 64% 63% 65% 66% Median PFS7 (95% CI) NR (2.0, NE) NR (2.8, NE) 14.9 mo (7.2, NE) 14.8 mo (6.6, NR) 12-month PFS (95% CI) 51% (28, 70) 53% (34, 69) 54% (45.8, 60.7) 52% (35.8, 66.4) Median DoR8 (95% CI) NR (1.7, NE) NR (2.1, NE) 26.9 mo (13.6, NE) 12.6 (5.7, NR) 12L LBCL 4+ HLA matched, dosed with 80M vispa-cel CAR-T cells 22L (N=32) and 3L+ (N=3) LBCL patients treated with 40M, 80M, or 120M vispa-cel CAR-T cells optimized for multiple factors, including 2+ HLA matched and young donor 3Yescarta prescribing information 4Yescarta FDA Statistical Review 1Apr2022 5Breyanzi prescribing information 6Breyanzi FDA Statistical Review 24Jun2022 7Median follow up 6.0 mo for confirmatory; 11.8 mo for optimized; 22.1 mo for ZUMA-7; 6.2 mo for TRANSFORM 8Median follow up 5.1 mo for confirmatory; 7.9 mo for optimized; NR for ZUMA-7; 4.3 mo for TRANSFORM Note: For axi-cel and liso-cel, efficacy data is from respective study IRC (Independent Review Committee) CR: complete response; DoR: duration of response; HLA: human leukocyte antigen; NE: not evaluable; NR: not reached; mo: month; ORR: overall response rate; PFS: progression-free survival Efficacy data cutoff 29Sept2025 16 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. Vispa-cel efficacy on par with autologous CAR-T cell therapies in 4+ HLA matched confirmatory cohort 1 3 6 9 12 15 Confirmatory cohort1 (N=22) ORR = 18/22 (82%) CR rate = 14/22 (64%) DoCR = NR HLA match Donor age 4 O 5 Y 9 Y 5 Y 5 Y 7 Y 5 Y 4 Y 5 Y 8 Y 6 Y 6 Y 4 O 9 Y 5 Y 5 Y 4 Y 5 Y 7 Y 5 Y 6 Y 4 Y CR: complete response PR: partial response SD: stable disease PD: progressive disease Death 12L LBCL 4+ HLA matched, dosed with 80M vispa-cel CAR-T cells One vispa-cel-related grade 5 IEC-HS occurred on day 25 post-infusion Certain patients converted from CR or PR to PD at various assessments time points as indicated in the chart above Based on previously reported data from approved autologous CAR -T cell therapies; Caribou has not performed any comparative analysis directly with autologous CAR-T cell therapies (see Important Information) Data cutoff date 29Sept2025 CR: complete response; DoCR: duration of complete response; HLA: human leukocyte antigen; IEC-HS: immune effector cell-associated hemophagocytic lymphohistiocytosis-like syndrome; NR: not reached; O: old; ORR: overall response rate; Y: young Visit (months)

17 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. Optimized vispa-cel product profile drives deep, durable responses 1 6 12 18 21 243 9 15 Optimized vispa-cel profile (N=35)1 ORR = 30/35 (86%) CR rate = 22/35 (63%) DoCR = NR Dose HLA 40M 3L+ 2 80M 2L 2 80M 2L 3 120M 2L 4 120M 2L 5 80M 2L 3 80M 2L 5 80M 3L 6 80M 2L 2 80M 2L 2 80M 2L 9 80M 2L 5 80M 2L 5 80M 2L 7 40M 3L 2 80M 2L 5 80M 2L 4 80M 2L 5 80M 2L 8 80M 2L 6 80M 2L 6 120M 2L 3 120M 2L 2 80M 2L 9 80M 2L 2 80M 2L 5 80M 2L 5 80M 2L 4 80M 2L 5 120M 2L 6 80M 2L 7 80M 2L 3 80M 2L 5 80M 2L 6 80M 2L 4 CR: complete response PR: partial response SD: stable disease NA: not assessed PD: progressive disease Death Lost to follow-up Long-term follow-up // 36 48// Visit (months) 12L (N=32) and 3L+ (N=3) LBCL patients treated with 40M, 80M, or 120M vispa-cel CAR-T cells optimized for multiple factors, including 2+ HLA matched and young donor *Patient diagnosed with lung adenocarcinoma after D28 scan revealed a non-responsive lung nodule and was taken off study and enrolled on our long-term follow-up study. Patient last known to be in continued response without additional anti-lymphoma therapy at one year post vispa-cel Long-term follow-up data reflect the last known response; marked timepoints indicate confirmation of no disease progression One vispa-cel-related grade 5 IEC-HS occurred on day 25 post-infusion Certain patients converted from CR or PR to PD at various assessments time points as indicated in the chart above Data cutoff date 29Sept2025 CR: complete response; DoCR: duration of complete response; HLA: human leukocyte antigen; IEC-HS: immune effector cell-associated hemophagocytic lymphohistiocytosis-like syndrome; NR: not reached; ORR: overall response rate; RP2D: recommended phase 2 dose 18 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. Vispa-cel safety profile allows for outpatient administration and expansion to community sites Vispa-cel Axi-cel ZUMA-7 N=1703,4 Liso-cel TRANSFORM N=925,6All treated N=84 Confirmatory cohort N=221 Optimized profile N=352 All grade ≥Gr 3 All grade ≥Gr 3 All grade ≥Gr 3 All grade ≥Gr 3 All grade ≥Gr 3 Neurotoxicity,7 n (%) 12 (14) 4 (5) 1 (5) 0 (0) 1 (3) 0 (0) 124 (74)8 42 (25)8 11 (12) 4 (4) CRS, n (%) 46 (55) 1 (1) 13 (59) 1 (5) 19 (54) 1 (3) 157 (92) 11 (7)8 45 (49) 1 (1) Infections, n (%) 43 (51) 21 (25) 9 (41) 4 (18) 20 (57) 6 (17) N/R (41) N/R (14) N/R 14 (15) Prolonged cytopenias9 NA 22/80 (28) NA 5/19 (26) NA 7/32 (22) NA 49 (29) NA 40 (43) IEC-HS, n (%)10 2 (2) 2 (2) 1 (5) 1 (5) 1 (3) 1 (3) NR NR 1 (1) 0 (0) FOR ILLUSTRATIVE PURPOSES ONLY. No head-to-head trials between these products have been conducted. Caution is advised when comparing results of different clinical studies as there are differences in patient populations, follow-up times, clinical trial phases, subject characteristics, trial design, and other factors. See Important Information. 12L LBCL 4+ HLA matched, dosed with 80M vispa-cel CAR-T cells 22L (N=32) and 3L+ (N=3) LBCL patients treated with 40M, 80M, or 120M vispa-cel CAR-T cells optimized for multiple factors, including 2+ HLA matched and young donor 3Locke et al; NEJM 2022; 4Yescarta prescribing information 5Abramson et al; BLOOD 2023, 6Breyanzi FDA statistical review 7Vispa-cel includes: ICANS; ZUMA-7 includes: all neurologic events; TRANSFORM includes: liso-cel–related investigator-identified events. Note: ICANS was formally defined in 2018 (ASTCT consensus), limiting comparability across studies 8N=168 9For vispa-cel, prolonged cytopenias are defined as Grade 3 or 4 neutropenia, thrombocytopenia, or anemia ongoing at day 28 (+/- 5 days) post CAR-T infusion, based on laboratory data, distinct from investigator-reported clinical adverse events. Analysis includes patients with assessments at day 28 (+/-5 days). Prolonged cytopenia for ZUMA-7 defined as ongoing at 30 days post axi-cel. Prolonged cytopenia for TRANSFORM defined as ongoing 35 days post liso-cel. 10One vispa-cel-related grade 5 IEC-HS that occurred day 25 post-infusion. IEC-HS was formally characterized in 2023 (ASTCT consensus) and previously characterized broadly as HLH/MAS, limiting comparability across studies. HLH/MAS rates were not reported in ZUMA-7 CRS: cytokine release syndrome; GvHD: graft versus host disease; IEC-HS: immune effector cell-associated hemophagocytic lymphohistiocytosis-like syndrome; NA: not applicable; N/R: not reported Data cutoff date 02Sept2025 19 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. Expansion and persistence correlate with duration of response in confirmatory cohort and with optimized vispa-cel profile Optimized profile 31 of 35 evaluable2 Confirmatory cohort 19 of 22 evaluable1 Response ≥3 months Progression <3 months Average of log transformed values shown; error bars represent standard error 1Progression <3 mo N=9; response ongoing ≥3 months N=10 2Progression <3 mo N=13; response ongoing ≥3 months N=18; ongoing response <3 mo N=1 (not shown) Response categories determined at Month 3/Week 12 visit Data cutoff date 09Sept2025 20 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. Rapid hematologic and immunologic recovery after vispa-cel contributes to generally well-tolerated safety profile B cell, T cell, and NK cell depletion and recovery to baseline levels LLOQ B cells LD T cells NK cells B cells Absolute neutrophil and platelet count recovery to Grade ≤2 N=84; average of log transformed values shown with ribbons reflecting standard error Baseline B cells absolute levels calculated with data points ≥ LLOQ LD: lymphodepletion; LLOQ: lower limit of quantification Data cutoff for T, B, and NK cell recovery is 09Sept2025; data cutoff for neutrophil and platelet count recovery is 02Sept2025 T and NK cells recover in ~4 weeks B cells recover in ~300 days 95% of neutropenia and 86% of thrombocytopenia recover by 60 days Pr ob ab ili ty o f r ec ov er y Time (Days) 0 30 60 90 Neutropenia 82 10 3 1 Thrombocytopenia 73 12 8 2 1.0 0.8 0.6 0.4 0.2 0.0

21 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. Vispa-cel safety, efficacy, and durability demonstrate potential as best-in-class allogeneic CAR-T cell therapy for r/r LBCL Data cutoff 02Sept2025 for safety and 29Sept2025 for efficacy Based on previously reported data from approved autologous CAR-T cell therapies; Caribou has not performed any comparative analysis directly with autologous CAR-T cell therapies (see Important Information) CR: complete response; ORR: overall response rates; PFS: progression-free survival; r/r: relapsed or refractory Generally well-tolerated safety profile that enables utilization of vispa-cel outpatient and in the community setting Data show vispa-cel has the potential to be the best-in-class allogeneic CAR-T cell therapy for large B cell lymphoma patients Efficacy in confirmatory cohort demonstrate vispa-cel is on par with autologous CAR-T cell therapies 86% ORR, 63% CR, 53% PFS at 12 months with optimized vispa-cel product demonstrates efficacy and durability are on par with autologous CAR-T cell therapies 22 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. Pivotal approach: randomized, controlled pivotal phase 3 trial to support full approval in 2L LBCL 1LBCL subtypes include DLBCL NOS, HGBL (with MYC and BCL2 and/or BCL6), transformed DLBCL from FL or MZL, FL3B, PMBCL 2Single infusion of vispa-cel following a lymphodepletion regimen of cyclophosphamide 60 mg/kg/d x 2d and fludarabine 25 mg/m2/d x 5d 2L: second-line; CRR: complete response rate; DLBCL: diffuse large B cell lymphoma; DoCR: duration of complete response; DoR: duration of response; FL3B: follicular lymphoma grade 3B; HGBL: high-grade B cell lymphoma; LBCL: large B cell lymphoma; NOS: not otherwise specified; MZL: marginal zone lymphoma; ORR: overall response rate; OS: overall survival; PFS: progression -free survival; PMBCL: primary mediastinal large B cell lymphoma; QOL: quality of life Vispa-cel Inclusion criteria: • 2L LBCL1 • CD19 naïve • Transplant- and autologous CAR-T- ineligibleSc re en in g Primary endpoint: PFS, interim analysis planned Secondary endpoints: ORR, CRR, DoR, DoCR, OS, QOL, safety 80x106 CAR-T cells following LD2 Investigator choice of standard of care immunochemotherapy regimens N ~2 50 Pivotal trial design based on internal analysis and FDA interactions to date; Company plans further engagement with FDA to refine pivotal trial protocol prior to initiation of pivotal trial 23 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. Vispa-cel safety profile and off-the-shelf availability aims to bridge the care gap Vispa-cel is designed for patients who cannot wait for treatment Safety profile allows for outpatient use at new sites of care Patients who live 2 to 4 hours from a treatment center are ~40% less likely to receive CAR-T1 1Ahmed N et al. Transplantation and Cellular Therapy 30: 714-725, 2024 Reduced logistical burden; no apheresis, vispa-cel manufactured from healthy donors Bringing vispa-cel closer to where patients live by leveraging academic and community hospitals Illustration purposes only, does not represent actual sites for the pivotal trial Academic Community 24 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. Meaningful commercial opportunity in 2L LBCL ~29,000 1L DLBCL patients treated in U.S.1 Only 1 in 4 2L patients receive autologous CAR-T cell therapies2 ~10,000 2L DLBCL patients treated in the U.S.1 ~$6.6B ~$13B Total LBCL global market growth from 2025 to 20331 1DRG 2025 model on file; numbers are estimate for DLBCL, does not include subtypes PMBCL, HGBL, grade 3B FL, and tMZL; global market estimated for G7: US, France, Italy, Spain, Germany, UK, and Japan 2Perales, M-A, et al. Poster 549, 2025 Tandem Meetings Even modest penetration leads to meaningful market opportunity for vispa-cel 2025 2033

25 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. Vispa-cel’s commercial-ready manufacturing enables orders of magnitude lower investment than auto CAR-Ts Potential for 96% lower COGS than current autologous CAR-T cell therapies COGS: cost of goods sold Small footprint at CDMO Easy and fast to expand Single 500 ft2 suite = 9,000 doses/yr Commercial-ready outputs Projected yield for 200-300 doses per batch Efficiency and flexibility Suite usable for any Caribou product On-demand starting materials 26 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. Vispa-cel drives deep, durable responses, demonstrating best-in-class allogeneic CAR-T cell therapy potential for r/r LBCL 86% ORR 53% 12-month PFS 63% CR rate No GvHD or Gr 3+ ICANS, <5% grade 3+ CRS, and manageable rates of infections and prolonged cytopenias4 Efficacy and durability on par with autologous CAR-T cells1 Potential best-in-class allogeneic CAR-T cell therapy for safety, efficacy, and durability with optimized vispa-cel3 Pivotal trial in 2L LBCL Expected trial design2: randomized, controlled trial in CD19-naïve, auto CAR-T- and transplant-ineligible patients; control arm to be treated with investigator choice of standard of care immunochemotherapy regimens 1Based on previously reported data from approved autologous CAR-T cell therapies; Caribou has not performed any comparative analysis directly with autologous CAR-T cell therapies (see Important Information) 2Pivotal study approach based on interactions with the FDA to date; the Company intends to further refine the pivotal trial design through continued engagement with the FDA prior to initiation of pivotal trial 3N=35; CD19 naïve, LBCL patients treated with 40M, 80M, or 120M vispa-cel CAR-T cells optimized for multiple factors, including 2+HLA matched and young donor 4Prolonged cytopenias are defined as Grade 3 or 4 neutropenia, thrombocytopenia, or anemia ongoing at day 28 (+/- 5 days) post CAR-T infusion, based on laboratory data, distinct from investigator-reported clinical adverse events. 2L: second-line; CR: complete response; CRS: cytokine release syndrome; GvHD: graft versus host disease; ICANS: immune effector cell-associated neurotoxicity syndrome; ORR: overall response rate; PFS: progression-free survival; r/r: relapsed or refractory Efficacy data cutoff 29Sept2025; safety data cutoff 02Sept2025 27 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. CB-011 Allogeneic anti-BCMA CAR-T cell therapy for r/r multiple myeloma (MM) 28 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. CB-011: delivering on the allogeneic CAR-T cell therapy promise with high response rates and low rates of infection for patients with r/r multiple myeloma Bispecifics CB-011 Tr e a tm e n t b u rd e n E ff ic a c y In fe ct io n High response rates2 with single dose High rates with limited or no B cell recovery1 Repeat dosing until relapse Single-dose treatment Weekly or bi-weekly treatment required for durability Low rates of infection2 and rapid immune recovery 1Frerichs, KA, et al. Blood Adv. 2024 Jan 9;8(1):194-206; Jelinek T, et al. Blood 144 (2024) 1934-1936; Schreiber S, et al. Mol. Therapy 3 (9) 4130-4134; 2025 2Based on Grade 3+ infections, at recommended dose for expansion as seen in results from dose escalation portion of the CaMMouflage clinical trial Based on previously reported data from approved bispecific therapies; Caribou has not performed any comparative analysis directly with bispecifics (see Important Information) r/r: relapsed or refractory

29 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. CB-011: delivering on the allogeneic CAR-T cell therapy promise with broad access, rapid treatment, and scalability for r/r multiple myeloma 1Frerichs, KA, et al. Blood Adv. 2024 Jan 9;8(1):194-206; Jelinek T, et al. Blood 144 (2024) 1934-1936; Schreiber S, et al. Mol. Therapy 3 (9) 4130-4134; 2025 2Based on Grade 3+ infections, at recommended dose for expansion as seen in results from dose escalation portion of the CaMMouflage clinical trial 3Gilead Q3 2024 earnings call transcript; Poseida Therapeutics International Myeloma Society Meeting data call 2024 4Kourelis, T. et al. Transplant Cell Ther 2023 29(4):255-258 Based on previously reported data from approved bispecific and autologous CAR-T therapies; Caribou has not performed any comparative analysis directly with such therapies (see Important Information) MM: multiple myeloma; r/r: relapsed or refractory Bispecifics CB-011 1 dose per manufacturing batch Weeks to months from eligibility to lymphodepletion4 1 of 10 MM patients receive auto CAR-Ts3 E ff ic a c y In fe ct io n Potential for 50-100 doses per manufacturing batch at commercial launch No wait needed between eligibility and lymphodepletion Overcomes access challenges to treat more patients A cc e ss S p e e d S ca le Autologous CAR-T High response rates2 with single dose High rates with limited or no B cell recovery1 Repeat dosing until relapse Single-dose treatment Weekly or bi-weekly treatment required for durability Tr e a tm e n t b u rd e n Low rates of infection2 and rapid immune recovery 30 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. CB-011 drives deep, durable responses, potentially best-in-class allogeneic CAR-T cell therapy for patients with r/r MM CB-011 significantly outperforms bispecific response rates and safety1 1Based on previously reported data from approved bispecific therapies; Caribou has not performed any comparative analysis directly with bispecifics (see Important Information) 2RDE is 450x106 CAR-T cells with selected LD regimen of 500 mg/m2 cy and 30 mg/m2 flu daily x 3 days; N=12 BCMA-naïve patients treated with RDE CR: complete response; GvHD: graft versus host disease; IEC-EC: immune effector cell-associated enterocolitis; MM: multiple myeloma; MRD: minimal residual disease; neg: negative; ORR: overall response rate; PFS: progression-free survival; r/r: relapsed or refractory Data cutoff 24Sept2025 92% ORR (11/12) 91% MRD neg (10/11 evaluable patients) 75% ≥CR rate (9/12) 7 of 12 ≥VGPR at ≥6 months Manageable safety profile with no GvHD, IEC-EC, parkinsonism, or cranial nerve palsies observed at any dose level Recommended dose for expansion (RDE): 450x106 CAR-T cells2 Potentially best-in-class allogeneic CAR-T cell therapy Efficacy in BCMA-naïve patients at the RDE2: 31 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. Presenting CaMMouflage dose escalation data Disclosures – consulting or advisor role: Caribou, J&J, BMS, Karyopharm, and Sanofi Director of CAR-T and stem cell transplant clinical program at the center of excellence for multiple myeloma, and investigator for the CaMMouflage trial Mount Sinai Adriana Rossi, MD 32 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. 48 patients dosed with CB-011 in dose escalation for the CaMMouflage Phase 1 trial NCT05722418 1300 mg/m2 cy and 30 mg/m2 flu daily x 3 days 2500 mg/m2 cy and 30 mg/m2 flu daily x 3 days Data cutoff 24Sept2025 cy: cyclophosphamide; flu: fludarabine; HLA: human leukocyte antigen; IMiD: immunomodulatory drug; LD: lymphodepletion; M: million; MM: multiple myeloma; mAb: monoclonal antibody; MTD: maximum tolerated dose; PI: proteasome inhibitor; r/r: relapse or refractory; RP2D: recommended phase 2 dose; RDE: recommended dose for expansion Patients with r/r MM • ≥3 prior lines of therapy, including a PI, an IMiD, and an anti- CD38 antibody • Exclusions: prior CAR-T cell therapy and/or BCMA-targeted therapy within last 3 months Part A: 3+3 dose escalation • Objective: safety, determine MTD, RDE Part B: dose expansion • Objective: antitumor response, RP2D -5 to -3 DAYS DAY 0 28 DAYS 3 MONTHS 6 MONTHS 9 MONTHS 12 MONTHS Lymphodepletion1,2 Safety and tolerability Response assessment SINGLE DOSE of CB-011 Dose levels evaluated at 300 cy LD regimen1 50x106 CAR-T cells 150x106 CAR-T cells 450x106 CAR-T cells Dose levels evaluated at 500 cy LD regimen2 150x106 CAR-T cells 300x106 CAR-T cells 450x106 CAR-T cells 800x106 CAR-T cells Best HLA matching for class II implemented for 500 cy cohort

33 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. -5 to -3 DAYS DAY 0 28 DAYS 3 MONTHS 6 MONTHS 9 MONTHS 12 MONTHS Lymphodepletion1,2 Safety and tolerability Response assessment SINGLE DOSE of CB-011 Dose expansion to begin at 450M cell dose Patients with r/r MM • ≥3 prior lines of therapy, including a PI, an IMiD, and an anti- CD38 antibody • Exclusions: prior CAR-T cell therapy and/or BCMA-targeted therapy within last 3 months Part A: 3+3 dose escalation • Objective: safety, determine MTD, RDE Part B: dose expansion • Objective: antitumor response, RP2D Dose levels evaluated at 300 cy LD regimen1 50x106 CAR-T cells 150x106 CAR-T cells 450x106 CAR-T cells Dose levels evaluated at 500 cy LD regimen2 150x106 CAR-T cells 300x106 CAR-T cells 450x106 CAR-T cells 800x106 CAR-T cells Best HLA matching for class II implemented for 500 cy cohort NCT05722418 1300 mg/m2 cy and 30 mg/m2 flu daily x 3 days 2500 mg/m2 cy and 30 mg/m2 flu daily x 3 days Data cutoff 24Sept2025 cy: cyclophosphamide; flu: fludarabine; HLA: human leukocyte antigen; IMiD: immunomodulatory drug; LD: lymphodepletion; M: million; MM: multiple myeloma; mAb: monoclonal antibody; MTD: maximum tolerated dose; PI: proteasome inhibitor; r/r: relapse or refractory; RP2D: recommended phase 2 dose; RDE: recommended dose for expansion 34 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. Robust patient enrollment at CaMMouflage trial sites despite access to auto CAR-Ts and bispecifics 1Based on survey results from CaMMouflage investigators asking why patients were treated with CB-011 vs other treatment options r/r: relapsed or refractory; MM: multiple myeloma Key reasons why investigators chose CB-011 for their r/r MM patients1 One-and-done treatment with an off-the-shelf product No manufacturing wait time prevents clinical deterioration between apheresis and infusion Promising deep and durable responses observed as trial progressed LevineTennessee Oncology MCW Huntsman VCU UAB OHC Mt. Sinai Miami MSKCC Cleveland Clinic Hackensack UTSW U of Colorado Duke UK 16 of 16 trial sites have access to bispecifics and approved auto CAR-T cell therapies 35 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. High-risk, heavily pretreated patients enrolled in CaMMouflage dose escalation 1All patients treated with a single dose of CB-011 and a lymphodepletion regimen of either 500 mg/m2 cy or 300 mg/m2 cy with 30 mg/m2 flu daily x 3 days 2High-risk cytogenetics include t(4;14), del(17/17p), t(14;16), t(14;20), and amplification/gain (1q) 3EMD defined as: soft tissue plasmacytoma noncontiguous with bone or lytic lesion with paramedullary extension 44 patients received belantamab (ADC) one of whom also received elranatamab (bispecific), 3 patients received teclistamab (bispecific), and 1 patient received NK trispecific (CC-92329 (BCMAXNKG2D/CD16)) cy: cyclophosphamide; ECOG: Eastern Cooperative Oncology Group; flu: fludarabine; NK: natural killer; R-ISS: revised international staging system Data cutoff 24Sept2025 Patient and disease characteristics All patients1 (N=48) Age, years, median (range) 68.5 (49-84) Male, n (%) 33 (68.8) ECOG performance status, n (%) 0 13 (27) 1 35 (73) R-ISS disease stage, n (%) at diagnosis I 6 (13) II 17 (35) III 12 (25) Unknown 13 (27) High risk cytogenetics2, n (%) 27 (56) Extramedullary disease (EMD)3, n (%) 17 (35) Prior lines of therapy, median (range) 4 (3-11) Median time since diagnosis (years) 5.3 (1-15) Prior stem cell transplant, n (%) 30 (63) Prior exposure to BCMA therapy, n (%) 8 (17)4 36 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. Responses observed at all CB-011 doses with selected LD Selected LD 500 mg/m2 cy and 30 mg/m2 flu daily x 3 days all dose levels1 N=35 150M N=6 300M N=13 450M N=13 800M N=3 BCMA naïve N=4 BCMA naïve N=11 BCMA naïve N=12 BCMA naïve N=3 75% 55% 92% 33% 50% 18% 75% 0% 50% 27% 83% 0% 50% (2/4) 0% (0/3)4 91% (10/11)3 0% (0/3) 1150x106, 300x106, 450x106, or 800x106 CAR-T cells 2MRD negative at ≤10-5 3MRD not evaluable in one patient 4MRD available for 3 patients at the time of the data cut Data cutoff 24Sept2025 CR: complete response; cy: cyclophosphamide; flu: fludarabine; LD: lymphodepletion; M: million; MRD: minimal residual disease; ORR: overall response rates; sCR: stringent complete response; VGPR: very good partial response ORR CR/sCR ≥VGPR MRD- negativity rate2 All patients N=48

37 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. EMD (Y/N) MRD neg1 (Y/N) N Y Y Y N Y N Y Y Y Y Y N Y N Y N Y Y N Y Y N 12 mo9 mo6 mo3 moDay 28 One patient who had previously withdrawn from the trial died on day 90 of treatment-related ICAHT; one patient died of pneumonia on day 50 (not treatment related) Data shown are from BCMA-naïve patients dosed at 450M cell dose with LD regimen of 500 mg/m2 cy and 30 mg/m2 flu daily x 3 days Median follow up for all patients in 450M cohort is 8.3 months 1MRD negative at ≤10-5 Data cutoff 24Sept2025 CR: complete response; cy: cyclophosphamide; flu: fludarabine; ICAHT: immune effector cell-associated hematotoxicity; LD: lymphodepletion; M: million; mo: month; MRD: minimal residual disease; ORR: overall response rate; PD: progressive disease; PR: partial response; PFS: progression-free survival 92% ORR (11/12) sCR: stringent CR CR: complete response VGPR: very good PR PR: partial response PD: progressive disease Withdrawal from trial Death 91% MRD neg (10/11 evaluable patients) 75% ≥CR rate (9/12) 15 mo 7 of 12 ≥VGPR at ≥6 months CB-011 induces deep, durable responses at 450M in BCMA-naïve patients 38 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. All at selected LD1 (N=35) BCMA-naïve 450M at selected LD1 (N=12) Adverse events Any grade n (%) Grade ≥3 n (%) Any grade n (%) Grade ≥3 n (%) Infections 17 (49) 5 (14) 8 (67) 3 (25) CRS 11 (31) 1 (3) 4 (33) 1 (8) ICANS 3 (9) -- 3 (25) -- IEC-HS 3 (9) 1 (3) 1 (8) 1 (8) IEC-EC -- -- -- -- GvHD -- -- -- -- Prolonged cytopenias2 NA 11/33 (33) NA 5/12 (42) 1LD regimen of 500 mg/m2 cy and 30 mg/m2 flu daily x 3 days 2Any continued ≥ grade 3 cytopenia based on laboratory data at ≥ day 35; denominator is those evaluable at day 35 (+/ -5 days) Data cutoff 24Sept2025 AE: adverse event; CRS: cytokine release syndrome; cy: cyclophosphamide; flu: fludarabine; GvHD: graft-versus-host disease; ICANS: immune effector cell-associated neurotoxicity syndrome; ICAHT: immune effector cell-associated hematotoxicity; IEC-HS: immune effector cell-associated hemophagocytic lymphohistiocytosis-like syndrome; IEC-EC: immune effector cell-associated enterocolitis; LD: lymphodepletion; NA: not applicable • Notable AEs are manageable • 3 events of note at 450M dose level: • 1 grade 5 ICAHT (CB-011-related) on day 90 • 1 grade 5 pneumonia (unrelated to CB-011) on day 50 • 1 grade 4 Guillain-Barré Syndrome (CB-011-related) on day 129, resolving • 1 event of note at 300M dose level: • 1 grade 5 RSV (unrelated to CB-011) on day 73 • Prophylactic measures for cytopenias and infections and early intervention for IEC-HS have been successfully implemented in the protocol No GvHD, IEC-EC, parkinsonism, or cranial nerve palsies observed at any dose level 39 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. CB-011 expansion is correlated with responses at selected LD Flow PK data up to Day 28 visit from 35 patients treated with 500 mg/m2 cy and 30 mg/m2 flu daily x 3 days; arithmetic mean with SE shown Data cutoff 24Sept2025 AUC: area under the curve; Cmax: maximum concentration; Tmax: time to maximum concentration; LD: lymphodepletion; PK: pharmacokinetic; SE: standard error Responders (N=23) Non-responders (N=12) Days A bs ol ut e nu m be r o f C A R+ c el ls /µ l Measurement Responders Non-responders Tmax (median) Day 14 Day 8 Cmax (mean) 104 cells/µl 1 cells/µl AUC (mean) 0-week 4 976 cells/µl 4 cells/µl 40 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. Rapid recovery of endogenous T and NK cells contributes to the manageable safety profile Average of log transformed values shown with ribbons reflecting standard error; dotted line is lower limit of quantification (LLOQ) for B cells Data through week 6 for 34 patients who received CB-011 with LD regimen of 500 mg/m2 cy and 30 mg/m2 flu daily x 3 days 1van de Donk N et al. 21st IMS (Annual Meeting, Brazil), 2024, P-090 2Tabbara N, et al. Hematology Am Soc Hematol Educ Program. 2024;(1):116-125 Data cutoff 24Sept2025 NK: natural killer B cells T cells NK cells • Patient T cell depletion enables CAR-T cell expansion • Fast recovery of T and NK cells quickly reinstates the patient’s natural immunity, likely contributing to the favorable safety profile 1,2 • Balancing CAR-T cell activity and recovery of natural immune activity is a key differentiator for Caribou’s allogeneic CAR-T programs N=34 Days lo g1 0( AB S+ 1)

41 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. Single-dose treatment with rapid immune reconstitution differentiates CB-011 as an off-the-shelf option CB-011 offers single-dose approach with high response rates and manageable safety profile Allogeneic CAR-T cells Time • Repeat dosing Bispecifics • Limited or no B cell recovery • Single dose • Rapid immune cell reconstitution 42 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. CB-011 has the potential to shift the paradigm for treatment of patients with r/r multiple myeloma Manageable safety profile with no GvHD, colitis, parkinsonism, or cranial nerve palsies observed Potential to expand access and bring meaningful benefit to patients who urgently need a readily available, single-dose option Deep, durable responses observed in heavily pre-treated population 92% ORR, 75% ≥CR rate, 91% MRD negativity, and 7 of 12 ≥VGPR at ≥6 months at recommended dose for expansion MRD negative at ≤10-5 Data cutoff 24Sept2025 CR: complete response; GvHD: graft-versus-host disease; MRD: minimal residual disease; ORR: overall response rate; r/r: relapsed or refractory 43 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. Discussion with lymphoma and multiple myeloma clinicians Professor of hematology/oncology and division director for hematologic malignancies and cellular therapeutics University of Kansas Cancer Center Joseph P. McGuirk, DOTina Albertson, MD, PhD Chief medical officer Caribou Biosciences Disclosures for Dr. McGuirk AlloVir: Consultancy, Honoraria, Research Funding, Speakers Bureau; Juno Therapeutics: Consultancy, Honoraria, Research Funding; Magenta Therapeutics: Consultancy, Honoraria, Research Funding; Kite, a Gilead Company: Consultancy, Honoraria, Research Funding, Speakers Bureau; Nektar: Consultancy, Honoraria; BMS: Consultancy, Honoraria, Speakers Bureau; Orca Bio: Research Funding; Sana: Honoraria; CRISPR Therapeutics: Consultancy, Research Funding; In8bio, Inc.: Other: IIT Clinical Trial; Sanofi: Honoraria, Consultancy, Speakers Bureau; Autolus: Consultancy, Honoraria, Research Funding; Astellas Pharma: Consultancy, Research Funding; Gamida Cell: Consultancy, Honoraria, Research Funding; Caribou: Consultancy, Honoraria, Research Funding; Incyte: Consultancy, Honoraria Director of CAR-T and stem cell transplant clinical program at the center of excellence for multiple myeloma Mount Sinai Adriana Rossi, MD 44 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. Caribou is defining a new era in CAR-T cell therapy with two leading allogeneic programs ~$159M cash, cash equivalents, and marketable securities1 Continued engagement with FDA to refine pivotal trial prior to initiation Vispa-cel Dose expansion to initiate by EOY 2025 with data in 2026CB-011 1Preliminarily, Caribou expects to report $159.2M in cash, cash equivalents, and marketable securities as of September 30, 2025. Caribou is currently finalizing its financial results for the three and nine months ended September 30, 2025. While complete financial information is not yet available, the results presented above reflect preliminary estimates and the financial closing procedures for the quarter end are not yet complete. As a result, final results may vary from these preliminary estimates. See Important Information.

45 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. 1 Initial safety and efficacy data on the confirmatory cohort (20 patients) with partial HLA matching, with a minimum of six months of follow up for the majority of patients, as well as an update on the larger, maturing dataset presented previously. Q&A 46 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. Gratitude for patients, caregivers, and investigators UC Irvine UC San Diego HonorHealth Huntsman Baylor MD Anderson OHC Atlantic HealthUniv of Iowa University of AZ Augusta Univ Arkansas Univ of Kentucky MCW BMTGA Montefiore Advent Health Norton Hackensack USC Univ of Penn NYU Langone Banner MD Anderson Vanderbilt Tufts OHSU VA Commonwealth VA Oncology Assoc. Israel - ANTLER Hadassah University Hospital Rabin Medical Center Tel Aviv Sourasky Medical Center The Sheba Fund for Health Services and Research Australia - ANTLER Epworth Hospital Royal Perth Westmead Hospital Miami MSKCC Cleveland Clinic UTSW U of Colorado Duke UAB Tennessee Oncology ANTLER trial sites CaMMouflage trial sites Appendix Vispa-cel appendix slides

49 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. Vispa-cel is generally well tolerated Treatment-emergent adverse events (TEAEs1) in ≥25% of all patients 1TEAEs are defined as adverse events (AEs) with a start date on or after the vispa-cel infusion date 22L LBCL 4+ HLA matched, dosed with 80M vispa-cel CAR-T cells 32L LBCL patients treated with 80M vispa-cel CAR-T cells optimized for multiple factors, including 2+HLA matched and young donor 42L (N=32) and 3L+ (N=3) LBCL patients treated with 40M, 80M, and 120M vispa-cel CAR-T cells optimized for multiple factors, including 2+HLA matched and young donor CRS: cytokine release syndrome; TEAE: treatment-emergent adverse event Data cutoff date 02Sept2025 Event, n (%) All treated (N=84) Confirmatory cohort2 (N=22) Optimized profile at RP2D3 (N=27) Optimized profile4 (N=35) Any grade ≥Grade 3 Any grade ≥Grade 3 Any grade ≥Grade 3 Any grade ≥Grade 3 Any TEAE 83 (99) 73 (87) 21 (95) 16 (73) 26 (96) 20 (74) 34 (97) 26 (74) Thrombocytopenia 52 (62) 50 (60) 10 (45) 10 (45) 14 (52) 14 (52) 19 (54) 19 (54) CRS 46 (55) 1 (1) 13 (59) 1 (5) 17 (63) 1 (4) 19 (54) 1 (3) Anemia 44 (52) 36 (43) 7 (32) 5 (23) 9 (33) 5 (19) 12 (34) 8 (23) Neutropenia 33 (39) 30 (36) 5 (23) 5 (23) 6 (22) 6 (22) 9 (26) 9 (26) Hypokalemia 22 (26) 1 (1) 7 (32) 1 (5) 6 (22) 1 (4) 8 (23) 1 (3) Leukopenia 22 (26) 21 (25) 1 (5) 1 (5) 2 (7) 2 (7) 3 (9) 3 (9) 50 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. Median onset and duration for neurotoxicity and CRS in vispa-cel compared to autologous CAR-T cells Vispa-cel Axi-cel ZUMA-7 N=1704,5 Liso-cel TRANSFORM N=926,7All treated N=84 Confirmatory cohort N=221 Optimized profile at 80M N=272 Optimized profile N=353 All grade All grade All grade All grade All grade All grade Neurotoxicity,8 n (%) 12 (14) 1 (5) 1 (4) 1 (3) 124 (74)9 11 (12)10 Median onset, days (range) 8.0 (6-34) 14.0 (14-14) 14.0 (14-14) 14.0 (14-14) 5.0 (1-133) 11.0 (7-17) Median duration, days (range) 2.0 (1-27) 1.0 (1-1) 1.0 (1-1) 1.0 (1-1) 15.0 (N/R) 4.5 (1-30) CRS, n (%) 46 (55) 13 (59) 17 (63) 19 (54) 157 (92) 45 (49) Median onset, days (range) 3.0 (0-22) 3.0 (0-14) 4.0 (0-14) 4.0 (0-14) 3.0 (1-10) 5.0 (1-63) Median duration, days (range) 3.0 (1-20) 3.0 (1-11) 3.0 (1-11) 2.5 (1-11) 7.0 (2-43) 4.0 (1-16) 12L LBCL 4+ HLA matched, dosed with 80M vispa-cel CAR-T cells 22L LBCL patients treated with 80M vispa-cel CAR-T cells optimized for multiple factors, including 2+HLA matched and young donor 32L (N=32) and 3L+ (N=3) LBCL patients treated with 40M, 80M, or 120M vispa-cel CAR-T cells optimized for multiple factors, including 2+HLA matched and young donor 4Locke et al; NEJM 2022; 5Yescarta prescribing information, 6Abramson et al; BLOOD 2023, 7Breyanzi FDA statistical review 8Vispa-cel includes ICANS; ZUMA-7 includes: all neurologic events; TRANSFORM includes: liso-cel–related neuro events. Note: ICANS was formally defined in 2018 (ASTCT consensus criteria), limiting comparability across studies 9N=168 10Investigator-identified neurologic toxicity CRS: cytokine release syndrome; N/R: not reported Data cutoff date for safety 02Sept2025 51 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. Vispa-cel ANTLER response duration for 2L patients who received optimized product at RP2D 1 6 12 18 21 243 9 15 HLA 2 3 3 5 2 2 9 5 5 7 5 4 5 8 6 6 9 2 5 5 4 5 7 3 5 6 4 *Patient diagnosed with lung adenocarcinoma after D28 scan revealed a non-responsive lung nodule and was taken off study and enrolled on our long-term follow-up study. Patient last known to be in continued response without additional anti-lymphoma therapy at one year post vispa-cel Long-term follow-up data reflect the last known response; marked timepoints indicate confirmation of no disease progression One vispa-cel-related grade 5 IEC-HS occurred on day 25 post-infusion Certain patients converted from CR or PR to PD at various assessments time points as indicated in the chart above Data cutoff date 29Sept2025 CR: complete response; HLA: human leukocyte antigen; IEC-HS: immune effector cell-associated hemophagocytic lymphohistiocytosis-like syndrome; ORR: overall response rate; RP2D: recommended phase 2 dose CR: complete response PR: partial response SD: stable disease NA: not assessed PD: progressive disease Death Lost to follow-up Long-term follow-up // 36 48// Visit (months) Optimized profile at RP2D (80M) (N=27) ORR = 22/27 (82%) CR rate = 17/27 (63%) 52 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. Baseline characteristics for patients enrolled in ANTLER Patient and disease characteristics All patients N=84 Confirmatory cohort1 N=22 Optimized profile at RP2D2 N=27 Optimized profile3 N=35 Age, years, median (range) 66 (20-86) 61 (20-83) 66 (20 -86) 63 (20-86) Male, n (%) 64 (76) 16 (73) 20 (74) 25 (71) ECOG, n (%) 0 40 (48) 13 (59) 14 (52) 19 (54) 1 44 (52) 9 (41) 13 (48) 16 (46) NHL subtype, n (%) DLBCL, NOS 48 (57) 14 (64) 16 (59) 21 (60) HGBL 13 (15) 4 (18) 5 (19) 5 (14) tFL 14 (17) 4 (18) 5 (19) 7 (20) tMZL 1 (1) - 1 (4) 1 (3) PMBCL 2 (2) - - 1 (3) MCL 3 (4) - - FL 2 (2) - - - MZL 1 (1) - - - Age adjusted IPI (%) 0-1 33 (39) 13 (59) 13 (48) 18 (51) 2 32 (38) 5 (23) 7 (26) 10 (29) 3 19 (23) 4 (18) 7 (26) 7 (20) Baseline LDH status (%) > ULN 33 (39) 10 (45) 13 (48) 16 (46) > 2x ULN 13 (15) 1 (5) 2 (7) 2 (6) Bulky disease4 17 (20) 2 (9) 3 (11) 4 (11) 12L LBCL 4+ HLA matched, dosed with 80M vispa-cel CAR-T cells 22L LBCL patients treated with 80M vispa-cel CAR-T cells optimized for multiple factors, including 2+HLA matched and young donor 32L (N=32) and 3L+ (N=3) LBCL 2+HLA matched and young donor, dosed with 40M, 80M, or 120M vispa-cel CAR-T cells 4Bulky disease defined by maximum baseline lesion diameter ≥7.5 cm Data cutoff 02Sept2025 2L: second-line; DLBCL, NOS: diffuse large B cell lymphoma, not otherwise specified; ECOG: Eastern Cooperative Oncology Group; HGBL: high grade B cell lymphoma; HLA: human leukocyte antigen; IPI: International Prognostic Index; LDH: lactate dehydrogenase; MCL: mantle cell lymphoma; MZL: marginal zone lymphoma; NHL: non-Hodgkin lymphoma; PMBCL: primary mediastinal B cell lymphoma; tFL: transformed follicular lymphoma; tMZL: transformed marginal zone lymphoma; ULN: upper limit of normal

53 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. Vispa-cel has a PD-1 KO designed to reduce CAR-T cell exhaustion Vispa-cel NHL cell 1 2 3 TCR KO MHC I Anti-CD19 CAR CD19 PD-L1 PD-1 KO Armored with 3 genome edits 1 2 3 TRAC gene knockout (KO) • Eliminates TCR expression, reduces GvHD risk Anti-CD19 CAR site-specific insertion into TRAC locus • Eliminates random integration, targets tumor antigen PD-1 KO for enhanced antitumor activity • Reduces CAR-T cell exhaustion • Potentially contributes to initial tumor debulking CAR: chimeric antigen receptor; KO: knockout; CD: cluster of differentiation; chRDNA: CRISPR hybrid RNA-DNA; CRISPR: clustered regularly interspaced short palindromic repeats; PD-1: programmed cell death protein 1; TCR: T cell receptor; TRAC: T cell receptor alpha constant; scFv: single-chain variable fragment 1 To Caribou’s knowledge. 1st CAR-T in the clinic with checkpoint disruption via PD-1 KO1 Cas9 chRDNA editing for reduced off-target editing and enhanced genomic integrity Anti-CD19 scFv FMC63 with a 4-1BB costimulatory domain CB-011 appendix slides 55 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. 1LD regimen of 500 mg/m2 cy and 30 mg/m2 flu daily x 3 days 2Infections and infestations are by system organ class Data cutoff 24Sept2025 system presented cy: cyclophosphamide; flu: fludarabine; LD: lymphodepletion All at selected LD1 N=35 BCMA-naïve 450M at selected LD1 N=12 Adverse events Any grade n (%) Grade ≥3 n (%) Any grade n (%) Grade ≥3 n (%) Neutropenia/neutrophil count decreased 28 (80) 27 (77) 10 (83) 10 (83) Anemia 21 (60) 16 (46) 10 (83) 8 (67) Thrombocytopenia/platelet count decreased 17 (49) 10 (29) 7 (58) 4 (33) Infections2 17 (49) 5 (14) 8 (67) 3 (25) Dizziness 11 (31) -- 4 (33) -- Cytokine release syndrome 11 (31) 1 (3) 4 (33) 1 (8) Fatigue 11 (31) 4 (11) 3 (25) 2 (17) Leukopenia 10 (29) 10 (29) 1 (8) 1 (8) Decreased appetite 10 (29) 2 (6) 4 (33) -- Constipation 9 (26) -- 1 (8) -- Pyrexia 9 (26) -- 3 (25) -- CB-011 has a manageable safety profile Treatment emergent adverse events (TEAEs) in ≥25% of patients at selected LD 56 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. High-risk, heavily pretreated patients enrolled in CaMMouflage Patient and disease characteristics All patients1 (N=48) All patients at selected LD2 (N=35) BCMA-naïve 450M at selected LD2 (N=12) Age, years, median (range) 68.5 (49-84) 69 (53-82) 71 (53-80) Male, n (%) 33 (68.8) 24 (69) 9 (75) ECOG performance status, n (%) 0 12 (27) 7 (20) 3 (25) 1 35 (73) 28 (80) 9 (75) R-ISS disease stage, n (%) at diagnosis I 6 (13) 4 (11) - II 17 (35) 12 (34) 4 (33) III 12 (25) 8 (23) 5 (42) Unknown 13 (27) 11 (31) 3 (25) High risk cytogenetics3, n (%) 27 (56) 19 (54) 9 (75) Extramedullary disease (EMD), n (%)4 17 (35) 12 (34) 5 (42) Prior lines of therapy, median (range) 4 (3-11) 4 (3-11) 4 (3-5) Median time since diagnosis (years) 5.3 (1-14.9) 5.3 (1-14.9) 4.8 (1.7-14.5) Prior stem cell transplant, n (%) 30 (63) 20 (57) 5 (42) Prior exposure to BCMA therapy, n (%) 8 (17)5 5 (14)6 - 1All patients treated with a single dose of CB-011 and a lymphodepletion regimen of either 500 mg/m2 cy or 300 mg/m2 cy with 30 mg/m2 flu daily x 3 days 2LD regimen of 500 mg/m2 cy and flu 30 mg/m2 daily x 3 days 3High-risk cytogenetics include t(4;14), del(17/17p), t(14;16), t(14;20), and amplification/gain (1q) 4EMD defined as: soft tissue plasmacytoma noncontiguous with bone or lytic lesion with paramedullary extension 54 patients received belantamab (ADC) one of whom also received elranatamb (bispecific), 3 patients received teclistamab (bispecific), and 1 patient received NK trispecific (CC-92329 (BCMAXNKG2D/CD16)) 62 patients received belantamab (ADC) one of whom also received elranatamb (bispecific), 2 patient received teclistamab (bispecific), and 1 patient received NK trispecific (CC-92329 (BCMAXNKG2D/CD16)) Data cutoff 24Sept2025 Cy: cyclophosphamide; ECOG: Eastern Cooperative Oncology Group; flu: fludarabine; LD: lymphodepletion; NK: natural killer; R-ISS: revised international staging system

57 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. Low rates of Grade ≥3 cytopenias at Day 35 in CaMMouflage All patients (N=48) All at selected LD (N=35) 450M BCMA naïve at selected LD (N=12) Grade ≥3 at anytime n (%) Grade ≥3 at D35 n (%) Grade ≥3 at anytime n (%) Grade ≥3 at D35 n (%) Grade ≥3 at anytime n (%) Grade ≥3 at D35 n (%) Anemia 26/48 (54) 6/43 (14) 19/35 (54) 5/33 (15) 8/12 (67) 2/12 (17) Neutropenia 41/48 (85) 4/40 (10) 29/35 (83) 3/30 (10) 11/12 (92) 2/12 (17) Thrombocytopenia 17/48 (35) 12/37 (32) 13/35 (37) 8/27 (30) 5/12 (42) 3/10 (30) Table reflects cytopenias based on laboratory data; grade ≥ 3 at D3 defined as cytopenia of those evaluable at D35 (+/-5 days) Includes patients with documented recovery Data cutoff: 24Sept2025 58 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. 1Includes both LD regimens: 300 mg/m2 or 500 mg/m2 cy with 30 mg/m2 daily x 3 days 2LD regimen of 500 mg/m2 cy and 30 mg/m2 flu daily x 3 days 3Infections and infestations are presented by system organ class Data cutoff 24Sept2025 cy: cyclophosphamide; flu: fludarabine; LD: lymphodepletion All patients1 N=48 All at selected LD2 N=35 BCMA-naïve 450M at selected LD2 (N=12) Adverse events Any grade n (%) Grade ≥3 n (%) Any grade n (%) Grade ≥3 n (%) Any grade n (%) Grade ≥3 n (%) Neutropenia/neutrophil count decreased 38 (79) 37 (77) 28 (80) 27 (77) 10 (83) 10 (83) Anemia 31 (65) 23 (48) 21 (60) 16 (46) 10 (83) 8 (67) Thrombocytopenia/platelet count decreased 24 (50) 15 (31) 17 (49) 10 (29) 7 (58) 4 (33) Infections3 22 (46) 5 (10) 17 (49) 5 (14) 8 (67) 3 (25) Leukopenia 17 (35) 17 (35) 10 (29) 10 (29) 1 (8) 1 (8) Dizziness 14 (29) -- 11 (31) -- 4 (33) -- Cytokine release syndrome 14 (29) 1 (2) 11 (31) 1 (3) 4 (33) 1 (8) Fatigue 14 (29) 5 (10) 11 (31) 4 (11) 3 (25) 2 (17) Pyrexia 13 (27) -- 9 (26) -- 3 (25) -- Decreased appetite 12 (25) 2 (4) 10 (29) 2 (6) 4 (33) -- CB-011 has a manageable safety profile Treatment emergent adverse events (TEAEs) in ≥25% of all treated patients 59 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. All patients1 (N=48) All at selected LD2 (N=35) BCMA-naïve 450M at selected LD2 (N=12) Adverse events Any grade n (%) Grade ≥3 n (%) Any grade n (%) Grade ≥3 n (%) Any grade n (%) Grade ≥3 n (%) Infections 22 (46) 5 (10) 17 (49) 5 (14) 8 (67) 3 (25) CRS 14 (29) 1 (2) 11 (31) 1 (3) 4 (33) 1 (8) ICANS 3 (6) -- 3 (9) -- 3 (25) -- IEC-HS 3 (6) 1 (2) 3 (9) 1 (3) 1 (8) 1 (8) IEC-EC -- -- -- -- -- -- GvHD -- -- -- -- -- -- Prolonged cytopenias3 NA 13/37 (35) NA 11/33 (33) NA 5/12 (42) 1Includes both LD regimens: 300 mg/m2 cy or 500 mg/m2 cy and flu 30 mg/m2 daily x 3 days 2LD regimen of 500 mg/m2 cy and flu 30 mg/m2 daily x 3 days Data cutoff 24Sept2025 3Any continued ≥grade 3 cytopenia at ≥day 35, denominator shown as patients who are evaluable at day 35 (+/-5 days) AE: adverse event; CRS: cytokine release syndrome; cy: cyclophosphamide; flu: fludarabine; GvHD: graft versus host disease; I CANS: immune effector cell-associated neurotoxicity syndrome; ICAHT: immune effector cell-associated hematotoxicity; IEC-EC: immune effector cell-associated enterocolitis; IEC-HS: immune effector cell-associated hemophagocytic lymphohistiocytosis-like syndrome; LD: lymphodepletion regimen; NA: not applicable No GvHD, IEC-EC, parkinsonism, or cranial nerve palsies observed at any dose level in CaMMouflage 60 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. CB-011 has potential to address unmet need and capture portion of increasing market opportunity ~192,000 people with MM in U.S.1 Only 1 in 10 MM patients receive autologous CAR-T cell therapies2 1NCI’s SEER 2025; total people with MM in U.S. reported in 2022; newly diagnosed patients reported in 2025 2Gilead Q3 2024 earnings call transcript; Poseida Therapeutics International Myeloma Society Meeting data call 2024 3Morris and Webber. Nat Rev Drug Discov 2025 Apr;24(4):244-245; markets included are United States, France, Germany, Italy, Spain, UK, and Japan 4Harrison RP, et al. Cytotherapy, 2019; 21:224-233 MM: multiple myeloma 0 10 20 30 40 50 60 2024 2028 2033 28.3B M aj or m ar ke t s al es (U S$ b ill io ns ) 43.7B 52.1B Estimated sales of all MM therapeutics3 ~36,000 newly diagnosed in U.S. per year1 Potential COGS for CB-011 ~80% lower at launch than current autologous CAR-Ts4

61 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. CB-011 450M at selected LD, BCMA naïve N=12 TECVAYLI1 (teclistamab) MajesTEC-1 N=165 ELREXFIO2 (elranatamab) MagnetisMM-3 N=123 TALVEY 3 (talquetamab) MonumenTAL-1 (weekly/Q2W dosing) * (N=30/N=44) LYNOZYFIC4 (linvoseltamab) LINKER-MM1 N=117 Median follow up 8.3 mos 14.1 mos 14.7 mos 11.7/4.2 mos 14.3 mos Efficacy ORR 92% 63% 61% 70%/64% 71% CR and sCR 75% 40% 35% 23%/23% 50% ≥VGPR 83% 59% 56% 57%/52% 63% Median DoR NR 18.4 mo NR 10.2 mo/ 7.8 mo 29.4 mo MRD-negativity rate 91% (10/11) 27% 90% 69% 91% 450M (N=12) All dosed at selected LD (N=35) Safety CRS Gr 3+ 8% 3% 0.6% 0% 3%/0% 0.9% Any 33% 31% 72% 58% 77%/80% 46% Neurotox (ICANS) Gr 3+ 0% 0% 0.6% 0% 0%/0% 3% Any 25% 9% 15% 3% 10%/5% 8% Infections Gr 3-4 17% 9% 45% 40% 7%/7% 36% Gr 5 8% 6% (12 deaths due to COVID-19) 7% 0%/0% 3% CB-011 compared to approved bispecific antibody therapies 1N Engl J Med 2022;387:495-505 (MRD: 44 pts (of 165 total) at 10-5 sensitivity); excluded prior BCMA 2Nature Medicine volume 29, 2259–2267 (2023) (MRD: Evaluable pts (N=29) with CR or better); excluded prior BCMA 3N Engl J Med 2022;387:2232-2244 (MRD: Among 16 evaluable pts with CR or sCR across all cohorts); 6pts received prior BCMA-targeted bsAb or CAR-T therapies 4J Clin Oncol. 2024 Aug 1;42(22):2702-2712 (MRD evaluable pts (N=21) with CR or better); Linvo trial allowed BCMA ADC (belamaf) only 10 pts received belamaf in the study NR: not reached FOR ILLUSTRATIVE PURPOSES ONLY. No head-to-head trials between these products have been conducted. Caution is advised when comparing results of different clinical studies as there are differences in patient populations, follow-up times, clinical trial phases, subject characteristics, trial design, and other factors. See Important Information. 62 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. CB-011 compared to allogeneic and autologous anti-BCMA CAR-T cell therapies CB-011 (N=12 450M, 500 cy BCMA naïve) P-BCMA-ALLO11 (N=23, 750 cy Total arm C) * ABECMA3 (Ide-cel) KarMMa-3 N=254 (efficacy) CARVYKTI4 (Cilta-cel) CARTITUDE-4 N=208 (efficacy) Anito-cel5 iMMagine-1 (Ph2) N=86 (efficacy) Median follow up 8.3 mos <3.5 mos2 18.6 mos 15.9 mos 9.5 mos Efficacy ORR 92% 91% 71% 85% 97% CR and sCR 75% 22%2 39% 73.% 62% ≥VGPR 83% 48%2 60% 81% 81% Median DoR NR -- 14.8 mo NR -- MRD negativity rate 91% (10/11) -- 20% 88% 93% 450M (N=12) All dosed, 500 cy (N=35) N=250 (Infections) N=225 (CRS/ICANS) N=208 (Infections) N=176 (CRS/ICANS) N=98 (safety) Safety CRS Gr 3+ 8% 3% 0% 5% 1% 1% Any 33% 31% 39% 88% 76% 83% Neurotox (ICANS) Gr 3+ 0% 0% 0% 3% 3%** 1% Any 25% 9% 13% 15% 21%** 9% Infections Gr 3-4 17% 9% 17% 24% 27% 10% Gr 5 8% 6% 0% 4% (7 deaths due to COVID-19) (1 death due to fungal infection) 1ASH Poster 2024; includes BCMA naïve and BCMA exposed patients; *the BCMA Naïve cohort in this study (N=9) had a reported ORR of 100%, no other values were reported; if estimated visually from the graph presented, CR and sCR are estimated to be ~20% and VGPR is estimated to be ~55% (see Important Information) 2Poseida IMS Investor Presentation 2024 3N Engl J Med 2023;388:1002-1014 (MRD: Within 3 mos before occurrence of at least a CR) 4N Engl J Med 2023;389:335-347 (MRD negativity at any time during trial; in 144 evaluable pts); ** Includes movement/neurocognitive AEs in addition to ICANS 5ASH 2024 oral presentation; Arcellx Corporate Presentation Feb 2025 (MRD: Of 58 evaluable pts at 10-5 sensitivity) NR: not reached FOR ILLUSTRATIVE PURPOSES ONLY. No head-to-head trials between these products have been conducted. Caution is advised when comparing results of different clinical studies as there are differences in patient populations, follow- up times, clinical trial phases, subject characteristics, trial design, and other factors. See Important Information. 63 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. Armored with 4 edits using chRDNA genome- editing technology for precision and significantly reduced off-target editing Removal of HLA class I surface expression mimics a 6 of 12 HLA match 63 CB-011: allogeneic anti-BCMA CAR-T cell therapy armored with immune cloaking to reduce T and NK cell rejection CB-011 B2M–HLA-E Anti-BCMA CAR TCR KO B2M KO Designed for functional persistence Immune cloaking implemented through: • B2M gene KO to slow down T cell-mediated rejection • B2M–HLA-E-peptide fusion insertion to blunt NK cell-mediated rejection B2M: beta-2 microglobulin; BCMA: B cell maturation antigen; HLA: human leukocyte antigen; KO: knockout; NK: natural killer; TCR: T cell receptor Corporate appendix slides

65 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. chRDNA guides promote on-target and reduce off-target edits First-generation all-RNA CRISPR-Cas chRDNA CRISPR hybrid RNA-DNA 66 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. Cas9 Cas12a chRDNA guides significantly improve editing specificity Editing assayed in primary human T cells Insert 1: 80% In se rt 2 : 7 6 % Knockout Knock-in Cas12a chRDNA genome editing + AAV6 transduction leads to >60% of manufacturing-scale engineered T cells with all 4 intended edits % e d it in g chRDNA guide on target chRDNA guide off targetAll-RNA guide on target All-RNA guide off target 0 2 4 6 20 40 60 80 100 all-RNA guide chRDNA PDCD1 ON OFF ON OFF TRAC B2M 0 2 4 6 20 40 60 80 100 chRDNA ON OFFON OFF % e d it in g 13.1 6.9 16.9 63.0 67 Caribou clinical data | November 3, 2025 ©2025 Caribou Biosciences, Inc. Engineering for improved activity against disease is key to unlocking the full potential of allogeneic cell therapies Caribou is implementing multiple armoring strategies Time C A R -T c e ll cy to to x ic a c ti v it y Checkpoint disruption (vispa-cel) PD-1 knockout (KO) sustains initial activity due to exhaustion resistance Immune cloaking (CB-011) B2M KO plus B2M–HLA-E fusion knock-in may delay host immune rejection Pretreatment lymphodepletion “Standard” allo CAR-T Checkpoint disruption Immune cloaking