#1/11 BenevolentAl Preliminary results for the year ended 31 December 2022 16 March 2023 BenevolentAl Proprietary ΑΙ Benevolent#2Disclaimer Forward-Looking Statements This document may contain forward-looking statements. Forward-looking statements are statements that are not historical facts and may be identified by words such as "plans", "targets", "aims", "believes", "expects", "anticipates", "intends", "estimates", "will", "may", "should" and similar expressions. Forward-looking statements include statements regarding objectives, goals, strategies, outlook and growth prospects; future plans, events or performance and potential for future growth; economic outlook and industry trends; developments in BenevolentAl's markets; the impact of regulatory initiatives; and/or the strength of BenevolentAl's competitors. These forward-looking statements reflect, at the time made, BenevolentAl's beliefs, intentions and current targets/aims. Forward-looking statements involve risks and uncertainties because they relate to events and depend on circumstances that may or may not occur in the future. The forward-looking statements in this release are based upon various assumptions based on, without limitation, management's examination of historical operating trends, data contained in BenevolentAl's records, and third-party data. Although BenevolentAl believes that these assumptions were reasonable when made, these assumptions are inherently subject to significant known and unknown risks, uncertainties, contingencies and other important factors which are difficult or impossible to predict and are beyond BenevolentAl's control. Forward-looking statements are not guarantees of future performance and such risks, uncertainties, contingencies and other important factors could cause the actual outcomes and the results of operations, financial condition and liquidity of BenevolentAl or the industry to differ materially from those results expressed or implied by such forward-looking statements. The forward-looking statements speak only as of the date of this release. No representation or warranty is made that any of these forward-looking statements or forecasts will come to pass or that any forecast result will be achieved. BenevolentAl Proprietary Benevolent 2#3AGENDA The Benevolent PlatformTM & Business Model Joanna Shields, CEO Preliminary Results for the year ended 31 December 2022 Nick Keher, CFO Pipeline review Anne Phelan, CSO Outlook Joanna Shields, CEO#42022: A year of solid progress and growth for BenevolentAl Enhanced the Benevolent Platform™, progressed our pipeline and delivered in commercial collaboration with AstraZeneca BEN-2293: Completed Phase Ila study and expect top-line data in Q1 2023 BEN-8744: Submitted CTA in Dec 2022, expect to initiate a Phase I clinical trial in H1 2023 BEN-28010: Declared as a clinical candidate in July 2022, with preparation for IND-enabling studies ongoing 3 year collaboration expansion with AstraZeneca's into 2 new disease areas 3 3 4 new in-house drug programmes generated using the Benevolent Platform™ FDA BenevolentAl Proprietary additional novel targets selected for AstraZeneca's portfolio in-house assets transitioned into lead optimisation approval of COVID-19 treatment (baricitinib) first identified by BenevolentAl Benevolent 4#5BenevolentAl is a leading Al-enabled drug discovery & development company with a platform-generated pipeline DISCOVERY Industry-leading Al platform uncovers novel biology across diseases and drug modalities Expansive knowledge graph with multimodal data foundations Disease and drug modality agnostic Generate predictions with the aim to increase the probability of success DEVELOPMENT Portfolio of first-in-class and best-in-class programs, discovered using our proprietary Al platform Discovery BEN-2293 - Atopic Dermatitis BEN-8744 - Ulcerative Colitis BEN-9160 ALS BEN- 28010 - Glioblastoma asset in Phase II Preclinical 1 3 15 Platform-generated drug programmes assets in pre-IND/CTA BenevolentAl Proprietary Clinical LO Benevolent 5#6Powerful data foundations generate a 360° view of disease biology Experiments Assay Data (Binding, Omics Comparison, CRISPR Screens) Clinical Trial OMICS Genes Proteins Isoforms Transcripts & Variants Biological Systems Cellular Component Molecular Function Biological Process Mechanism Pathways Literature Scientific Literature Patent Literature Regulatory Documents Aetiology Diseases Symptoms Molecules Organic Compounds Preclinical Candidates Approved Drugs Antibodies Other Biologics Pharmacology Pharmacokinetics Coma BenevolentAl Proprietary Multimodal approach combines over 85 data sources to provide a holistic view of disease biology Maximise our probability of clinical success by integrating disease traits, genetics and genomics data to generate endotype-specific target predictions Breaks down silos across therapeutic areas to connect shared mechanisms across disease Provides a proprietary integrated view of biomedical data that supports discovery and decision-making Benevolent 6#7The Benevolent Platform™ empowers scientists with industry-leading drug discovery Al Comprehensive data foundations DISEASE EXPLORATION Can we treat ALS by reversing Autophagy impairment in microglia by reducing oxidative stress" 000 Target Endpoint What we will measure Cell type Mechanism sign TARGET PREDICTIONS W B H TARGET IDENTIFICATION LARGE LANGUAGE MODEL GRAPH MODELS GENETIC & GENOMIC MODELS DATA QUERIES TRANSCRIPTOMICS Biology first MODELS TARGET ASSESSMENT Ⓒ O X Small Molecule XNOVELTY PATENT LANDSCAPE ( Alternative Modalities Hypothesis driven TRACTABILITY (?) SELECTIVITY BenevolentAl Proprietary TARGET VALIDATION Portfolio Entry ● ● Proprietary Al models reason across multi-modal data to discover novel targets Enables scientists to assess & select only the most promising targets to take into wet lab experiments Efficiently surfaces scientific evidence to support higher confidence decisions Data fed back into the Knowledge Graph to enhance future predictions Benevolent 7#8Our flexible business model unlocks multiple routes to value creation Al-Discovery Tools Knowledge Graph Precision Medicine 00 W Target Identification Molecular Design •To OWNED PIPELINE Platform-generated assets ● • In-house development LICENSING ● Platform-generated assets • Out-licensed at IND, end Phase I or end II PLATFORM COLLABORATIONS ● Economic Benefits • Platform Validation • Data generated enriches the Benevolent Platform™ NON-COMMERCIAL COLLABORATIONS ESG • Platform Validation • Data generated enriches the Benevolent Platform™ ● BenevolentAl Proprietary Benevolent 8#9Preliminary Results for the year ended 31 December 2022 Nick Keher, CFO#102022 Financial highlights Revenue R&D - Drug discovery ("DD") ¹ R&D - Product & technology ["P&T"]¹ G&A - Business operations ["Bus Ops"]¹ Underlying expenses related to share-based payments Other income Normalised operating loss Normalised EPS (in pence)² Weighted average ordinary shares outstanding (in millions) 1) Excludes exceptional costs related to the Business Combination 2) Normalised EPS also excludes taxation impact from exceptional items and finance income related to the Business Combination 31 December 2022 £'000 10,560 (43,179) (21,914) (16,500) (23,731) 166 (94,598) (72.6) 109.1 BenevolentAl Proprietary 2021 £'000 4,625 (27,129) (19,963) (13,944) (51,390) 90 (107,711) (104.6) 89.9 Revenue increase across AstraZeneca collaboration, with 3 additional novel targets in chronic kidney disease (x1) and idiopathic pulmonary fibrosis (x2). DD spend increase driven by advancing pipeline into later stages of development, particularly BEN-2293 adaptive Phase I/II and BEN-8744 CTA filing enablement in Dec-22. Net 4 named programmes added into pipeline during the year. P&T spend increase reflecting increased headcount, expected to plateau. Bus Ops spend +18%, driven predominantly by listing status and expected to maintain at this level. Benevolent 10#11Reported to Normalised¹ Operating loss Adjustments for: G&A - Direct Transaction costs G&A - Transaction-related listing service SBP expense G&A - Transaction-related employee-related SBP expense G&A - Impairment of assets G&A - Transaction-related stamp duty Revaluation of investments Normalised¹ group operating loss 1) Excludes exceptional costs related to the Business Combination 31 December 2022 £'000 (197,034) 11,255 83,067 3,883 3,740 491 (94,958) BenevolentAl Proprietary 2021 £'000 (121,322) 10,700 Reported operating loss significantly driven by non-recurring costs of Business Combination. payments also incurred from 2,911 Business Combination, and set to reduce beyond 2022. (107,711) Non-cash exceptional costs related to acceleration of share-based Non-cash listing service expenses reflects cost of share issuance to Odyssey SPAC as part of Business Combination net of assets acquired (before PIPE and backstop) Benevolent 11#12Cash flows focused upon drug and platform development Normalised¹ operating loss Depreciation & amortisation Foreign exchange Other employee-related SBP expense Cash flows from changes in working capital Cash expended from underlying operating activities Opening cash, cash equivalents and short-term deposits Closing cash, cash equivalents and short-term deposits 1) Excludes exceptional costs related to the Business Combination 31 December 2022 £'000 (94,598) 3,058 BenevolentAl Proprietary (3,141) 29,935 (13,094) (77,840) 40,553 130,182 £1.7m property-related leases. £3.2m gain from EUR holdings; £0.3m gain from USD holdings; £0.4m charge from operational. Non-Transaction-related equity awards removed from the P&L (no cash impact) Largely driven by increase in R&D tax credit receivable (£4.0m), as well as decrease in SBP employer-related tax provision (£6.2m). Year end cash position of £130.2m at the top end of our stated guidance. Benevolent 12#13Cash runway to Q4-2024 providing sufficient capital for key value inflection points Cash Runway Cash¹ at 31 December 2022 £130.2m 2022 cash burn of £64.7m before working capital movements Cash runway guidance assumes no future capital from licensing or collaboration agreements Multiple assets at, or close to, key value inflection points 1) Includes cash, cash equivalents and short-term deposits (maturity between 3 and 12 months). Benevolent 3 5 Capital allocation Fund Phase I trial for BEN-8744 in Ulcerative Colitis and commencement of Phase II trial in 2024 Fund CTA enabling work for BEN-28010 and begin Phase I trial in GBM in 2024 Progress balance of pipeline as well as fund CTA enabling work of novel ALS asset to complete in 2024 Continuous enhancement of the Benevolent Platform™ to support further collaborations Business Operations to support Group activities BenevolentAl Proprietary Benevolent 13#14Pipeline review Anne Phelan, CSO#15Robust pipeline entirely generated by the Benevolent Platform™ BEN-2293 | Atopic Dermatitis BEN-8744 | Ulcerative Colitis BEN-28010 | Glioblastoma Multiforme BEN-9160 | Amyotrophic Lateral Sclerosis Inflammatory Bowel Disease Amyotrophic Lateral Sclerosis Antiviral Oncology Oncology Parkinson's Disease Nonalcoholic Steatohepatitis (NASH) Oncology Parkinson's Disease Fibrosis Inflammation Chronic Kidney Disease Idiopathic Pulmonary Fibrosis Idiopathic Pulmonary Fibrosis Idiopathic Pulmonary Fibrosis Chronic Kidney Disease BAI pipeline as at end-December 2022 +10 Exploratory stage programmes Target ID Chemistry & Lead Opt AstraZeneca AstraZeneca AstraZeneca Preclinical AstraZeneca 27.01.2021 AstraZeneca 17.05.2022 15.12.2021 06.10.2022 06.10.2022 Phase I BenevolentAl Proprietary Phase II Highlights • All pipeline programmes generated using the Benevolent Platform™ • BEN-2293: Phase Ila read out Q1-2023 • BEN-8744: Novel target with zero prior linkage to UC. Delivered drug candidate within 2 years from programme initiation, starting Phase 1 H1-2023 • AstraZeneca: multi-year Target-ID collaboration expanded in Jan 2022 to include systemic lupus erythematosus & heart failure. Benevolent 15#16BEN-2293 for Atopic Dermatitis • Atopic dermatitis (AD) is the most common chronic inflammatory skin disease, characterized by intensely itchy, red, and swollen skin(¹) Affects 10-20% of children and up to 3% of adults (2) O Approximately 60-70% of all cases present with mild-moderate disease severity (3) O O Prevalence is rising (3), with market value in 7MM forecast to exceed $16.7 billion by 2030(4) • BEN-2293 is expected to treat atopic dermatitis by: inhibiting itch signaling and blocking nerve sensitization (TrkA) in addition to inhibiting Th1 /Th2-mediated dermal inflammation (TrkB, TrkC). . Using our Molecular Design expertise we were able to design a PanTrk inhibitor, equipotent against the 3 receptors Sources: (1) Weidinger et al. Nat Rev Dis Primers 2018; (2) GlobalData Report 2018: Atopic Dermatitis: Global Drug Forecast and Market Analysis to 2027; (3) Global Data Report 2018: Atopic Dermatitis: Epidemiology Forecast to 2027; (4) GlobalData Report 2022: Atopic Dermatitis: Global Drug Forecast and Market Analysis to 2030 TREATMENT FLOW BenevolentAl Proprietary Atopic Dermatitis Treatment Paradigm Mild Steroid Treatment Current Moderate Calcineurin inhibitors Future PDE4 inhibitors Topical JAKS BEN-2293 (Pan Trk) Other topicals (tapinarof) Severe Immuno- suppressants Dupilumab Other anti-IL13 mAbs Oral JAKS BEN-2293 will target Mild, Moderate and Severe AD patients, addressing unmet need as a steroid sparing alternative and in more severe patients undergoing treatment with systemics (e.g. dupilumab) that require add-on treatment 16#17BEN-2293 Phase lla clinical study designed to demonstrate efficacy in mild-moderate Atopic Dermatitis STUDY DESIGN: SAFETY & EFFICACY Recruitment completed end of 2022 91 patients randomised 1:1 for BEN-2293:Vehicle Patients (Inclusion Criteria) • Mild-Moderate AD o Baseline VIGA 2-3 o Baseline Itch NRS 24 o Atopic dermatitis affecting 1-30% BSA of treatable skin • 18-65 years Dosing 7 day medication washout followed by 3 day placebo run-in prior to dosing • 1% w/w BEN-2293, or placebo/vehicle applied twice daily to treatable lesioned skin up to 30% BSA • 28 days on treatment = Powered endpoints STUDY EFFICACY OUTCOMES Itch • Change from baseline in the Pruritus NRS (Worst itch over 24 hours and Current itch) Time to itch reduction ● • Fraction of patients achieving itch reduction Inflammation Change from baseline in EASI score Fraction of patients achieving improvement in EASI ● ● score Change from baseline in BSA affected by AD in treated area(s) • Change from baseline in VIGA Quality of life Change from baseline in POEM . Change from baseline in DLQI • Change from baseline in EQ-5D Change from baseline in PROMIS (sleep subscale) ● BenevolentAl Proprietary Benevolent 17#18BEN-2293 - Phase Ila study recruitment completed Screening Baseline assessments itch and AD score ● Wash out 7 day Wash out from any existing medication Current Status Re-baseline or exclude Run in 3 day Placebo run in Phase lla clinical study design LPLV - last patient last visit | DBL database lock | TLR - Top line results LPLV achieved end January 2023 91 patients randomised (65 moderate and 26 mild) DBL completed 1st March TLR anticipated end March Continuous safety monitoring 28d BID dosing over affected skin area 28 day BID dosing on affected skin upto a maximum of 30% BSA Placebo:Active (1:1), Moderate: Mild patient ratio 70:30, Total of 91 patients (50% active:50% placebo) BenevolentAl Proprietary Study endpoints - itch and AD rating scales Follow up 14 day Final safety assessment Our intention is to to partner this asset with a pharmaceutical company that has a focus in dermatology for continued clinical development and, if approved, commercialisation Benevolent 18#19BEN-8744 Ulcerative Colitis (UC) Affects 0.4% US population(¹), 1.7 million patients in 7MM(¹), forecast $7.8bn market by 2026(²) • A chronic, lifelong disease that causes inflammation and ulceration of the inner lining of the colon and rectum • 64% of patients are mild-moderate, 31% of patients are moderate-severe and 5% of patients are severe-fulminant • Efficacy - 20-40% of Moderate-severe patients do not respond to anti-TNF (main treatment paradigm) (3) • Safety - Treatments have many side effects - from steroids to anti-TNF and JAK inhibitors (black box warnings) (4) High unmet need for an alternative oral small molecule treatment option with improved safety profile and efficacy in treatment of refractory patients ● ● BEN-8744: Oral, peripherally-restricted, PDE10 inhibitor under development as a first-in-class treatment for refractory UC Phosphodiesterase 10 (PDE10) was identified by our TargetID platform as an entirely novel target for the treatment of UC/IBD Using our Molecular Design expertise we optimally designed a best in class peripherally restricted PDE10 inhibitor: BEN-8744 • BEN-8744 is expected to provide an efficacious disease modifying oral treatment for UC/IBD • BEN-8744 will target Moderate and Severe UC/IBD patients, meeting the unmet need left by existing therapies including: o Patients refractory to anti-TNFS or other biologics o Improved safety and tolerability profile compared to competitors o Aiming to use a Precision Medicine approach to target key responder patient cohorts and avoid the safety risks associated with ineffective therapies Sources: (1) GlobalData: Ulcerative Colitis, Global Drug Forecast and Market Analysis to 2026; (2) Evaluate Pharma: Gastro-intestinal, Inflammatory bowel disease (IBD), Ulcerative colitis, Worldwide Overview (report 17th Sep 2021); (3) Roda et al. Clin Transl Gastroenterol 2016; (4) Kobayashi et al Nat Rev Dis Primers 2020 and US FDA Drug Safety Communication 2021 Benevolent 19#20BEN-8744: potent activity in UC colon biopsies (w/6d)9-11 20000- 15000- 10000- 5000- Endoscopic Biopsy from UC patients DMSO Prednisolone Tofacitinib BEN-3218 Inflammatory cytokine measurement IL-8(pg/ml) 30000- 20000- Colonic mucosa organ culture and compound treatment 10000- DMSO Prednisolone Tofacitinib BEN-3218 Selective PDE10 inhibition showed comparable reduction of IL-6 and IL-8 to the positive controls in ex vivo UC biopsies Cross donor response summary: % inhibition cytokine release (mean % inhibition from N=15 IBD donors) 100- 1 XH 10 100 1000 1 BEN-8744 nM % inhibition IL6 % inhibition IL8 0+ 0.0001 0.001 0.01 100- 0+ 0.0001 0.001 0.1 jtttt 0.01 1 10 BEN-8744 nM 0.1 100 10000 1000 10000 Benevolent 20#21BEN-8744 results and progress to date Target validation 2019 2020 2021 Novel, potent advanced lead molecule developed within 2 years TARGET IDENTIFICATION Novel target for UC Discovered using Benevolent TargetID tools ✓ PDE10 has zero linkage to UC in all available biomedical literature ✓ Experimentally validated in ex-vivo UC colon samples from patients refractory to SoC treatment Candidate nominated CHEMISTRY Rapid and efficient lead optimisation ✓ Molecular Design tools enabled rapid and efficient lead optimisation ✔ Candidate nominated in Sep '21 Novel, potent, selective, peripherally restricted PDE10. Inhibitor, with low dose prediction Delivered drug candidate within 2 years from programme initiation BenevolentAl Proprietary 2022 2023 Preclinical development CTA Filed Phase I clinical study CLINICAL DEVELOPMENT Developing responder and progression endotypes We will develop responder and progression endotypes, adding molecular descriptors ✓ These will inform our trial design, patient selection and further target identification in UC ✓ Augmenting a further loop of iteration on an enriched graph Benevolent 21#22Outlook Joanna Shields, CEO#23"C "With the world's attention on Al applications that deliver real-world impact, we are strongly positioned to capitalise on this moment. With our substantial portfolio of platform-generated drugs, our work with big pharma and research collaborators, and our continued investment in state-of-the-art technology, we are showing every day how Al can be used to unlock the next wave of biopharma innovation." Joanna Shields, CEO Poised for growth: multiple value inflection points expected BEN-2293 • Top-line results of the Phase lla clinical trial expected in Q1 2023 • Out-license in H2 2023 subject to results BEN-8744 • Initiate the Phase I study in H1 2023 BEN-28010 Submit the CTA in H2 2023 2023 • Continued investment in state-of-the-art technology • Aim to sign additional commercial collaboration • Commence IND-enabling studies for at least one additional asset • Continue to progress pipeline assets to value inflection points Benevolent 23#24benevolent.com Thank you @benevolent_ai in benevolentai [email protected]#25Appendix#26World-class team We "build tech in the service of science" 48% Drug Discovery Bus Ops As of December 2022 15% 350+ World Class Scientists & Technologists 37% Tech Board of Industry Luminaries Combines deep expertise across Al, pharma, & drug discovery & development Baroness Joanna Shields CEO & Executive Director Jean Raby Non-Executive Director François Nader Chairman Jackie Hunter Non-Executive Director BenevolentAl Proprietary Susan Liautaud Non-Executive Director Nigel Shadbolt Non-Executive Director Olivier Brandicourt Non-Executive Director John Orloff Non-Executive Director Benevolent 26#27Proven to enhance drug discovery DISEASE-AGNOSTIC We can work on any therapeutic area due to the breadth and diversity of our data foundations. ACCELERATES DISCOVERY By combining our Al Platform, scientific expertise and wet lab facilities, we accelerate discovery and reduce discovery and development timelines. MODALITY-AGNOSTIC The Benevolent Platform™ can be applied to antibody and biologic targets, in addition to small molecule targets. IDENTIFIES NOVEL TARGETS Our predictive tools can surface targets that have never been considered for a disease before. BenevolentAl Proprietary BUILT FOR SCALE Our scalable and versatile Platform can support multiple in-house drug programmes and commercial collaborations. POTENTIAL TO INCREASE PROBABILITY OF SUCCESS By building higher confidence hypotheses in the earliest stages of drug discovery, we aim to reduce costly failures down the line. Benevolent 27#28Poised for growth: multiple value inflection points Internal pipeline depth and progression AZ Collaboration Other Platform Collaborations H1 2023 BEN-2293: AD Data expected Q1 2023 BEN-8744: UC Begin Phase I study H1 2023 H2 2023 BEN-2293: AD Out-licensing BEN-28010: GBM Submit Clinical Trial Application (CTA) Expect to add 4-6 names drug programmes to portfolio 2024 BEN-8744: UC Phase I data package 2024, Phase II to follow shortly after BEN-28010: GBM Initiate Phase I study Aim to progress 1-2 CTA/IND stage drug candidates every year BenevolentAl Proprietary Potential new targets added for SLE and Heart Failure - prior targets (3 for IPF and 2 for CKD) advancing Discussions with a number of parties ongoing Benevolent 28#29TrkC Atopic Dermatitis - BEN-2293, pan-Trk inhibition rationale ● ● NT3/TrkC potentiates stimulated Th2 T-cell inflammatory responses and synergistically enhances T-cell receptor induced IL-4 production by Th2 cells TrkB Mast cells within AD skin lesions express high levels of NT3 compared to normal controls AD Skin-resident eosinophils express elevated levels of TrkB (together with TrkA and C) and functionally respond to BDNF BDNF/TrkB inhibit eosinophil apoptosis and increase chemotactic index Healthy skin Stratum. Skin microbiota corneum Staphylococcus aureus. Stum granu sum Stratum spinosum Stratum basale Dermis Blood vessel B cell I cell Non-lesional skin IL-18 IL-33 TARC IL-25 MDC TSLP mm ILC2 cell CLA CCR4 Lichenification Barrier dysfunction, innate immune system activation and T2-driven inflammation and/or T.22-driven inflammation Keratinocyte Variable T1 and T, 17 activation IL-13 CCR10 CRTH2 T₁2 cell T,22 cell Allergen IL-4 IL-13 cell Acute lesional stage FCER1 Eosinophil OX40L ↓ H4R T,17 cell IgE IL-4 IL-13 IL-31 Trm cell IL-33 TSLP -Cutaneous sensory neuron BenevolentAl Proprietary Chronic lesional stage mu Langerhans cell Med Dermal dendritic cell IDEC ● TrkA TrkA levels in skin dramatically increase in response to inflammatory stimuli NGF produced by AD keratinocytes, is a major mediator of cutaneous hyperinnervation Increased NGF in the skin sensitizes primary afferents contributing to peripheral itch sensitization and chronic pruritus Involved in the inflammatory activation of mast cells and basophils Benevolent 29#30BEN-2293: Excellent skin penetration Experimental evidence supports high exposure in human skin at >IC90 free, and low exposure in blood with proposed clinical 1% ointment strength¹. 1% BEN-2293 ointment BID exceeds the exposure needed for PanTrk inhibition in both epidermis/upper dermis and lower dermis even at IC₂0¹ Human in vitro >>IC90 Minipig in vitro >> IC90 Minipig in vivo >IC90 Minipig in vivo Free plasma levels <<400 below IC50 Epidermis and upper dermis Lower Dermis Blood Healthy skin Stratum corneum Stratum granulosum Stratum spinosum Stratum basale Dermis Blood vessel Skin microbiota Staphylococcus aureus. B cell T cell Non-lesional skin IL-18 IL-33 TARC IL-25 MDC TSLP ILC2 cell CLA CCR4 IL-5 IL-13 Lichenification Barrier dysfunction, innate immune system activation and T 2-driven inflammation and/or T 22-driven inflammation Keratinocyte Variable T1 and T, 17 activation CCR10 CRTH2 T2 cell T22 cell Allergen IL-4 IL-13 cell Acute lesional stage FCER1 Eosinophil OX40L H4R T 17 cell IgE IL-4 IL-13 IL-31 Trm cell IL-33 TSLP Cutaneous sensory neuron Chronic lesional stage Langerhans cell Dermal dendritic cell CTA-enabling 28d Tox Package: Rat (IV) and Mini-pig (topical) = Safety margins > 20 fold for AUC and > 269 fold for Cmax to dose limited NOAELS¹ ¹(internal Company drug programme data) IDEC 30#31We expect BEN-2293 to provide both symptomatic relief and to reverse disease progression in atopic dermatitis • BEN-2293 is highly selective for Trk receptors, with IC50 potencies in the low nM range for TrkA, B, and C • BEN-2293 dose dependently inhibits release of inflammatory Th1 and Th2 cytokines TNFa, IFNY, IL-13, and IL-4 in human peripheral blood mononuclear cells (PBMCs) stimulated with an inflammatory T-cell stimulus (anti-CD3/CD28) • BEN-2293 inhibits the release of Calcitonin Gene-Related Peptide (CGRP), a mediator of itch, sensory nerve hypersensitisation and neurogenic inflammation, in dorsal root ganglion (DRG) isolated from adult rats and stimulated with NGF • BEN-2293 series significantly (p<0.05) reduced mouse ear inflammation following administration of PMA, significantly reducing expression of cytokines IL-1B, IL-4, IL-6, CXCL1, MCP-1, and Tarc • BEN-2293 demonstrates excellent tolerability and safety margins in IND/CTA-enabling toxicology studies Key: Tropomyosin-related kinase (Trk) receptor tyrosine kinase family, namely TrkA, TrkB, and TrkC; Nerve Growth Factor (NGF); Brain Derived Neurotrophic Factor (BDNF); Neurotrophin-3 (NTF-3)/NT3 BEN-2293 Inhibition of human primary T-cell activation BenevolentAl Proprietary Luminescence % Inhibition 25000 20000- 15000- 10000- 5000- 0.1 100- 75- 50- 25- 0- -25- -50- TNFa Inhibition of sensory neuron activation -3 -2 THE M 1000 10000 TIME TIITE 10 100 BEN-2293 nM 10 100 1000 10000 BEN-2293 nM Human PBMCs from 2 donors. Anti-CD3/CD28 stimulus +/- BEN-2293 Hillslope IC50 -1 0 Log BEN-2293 (μm) - Donor 1 Donor 2 2.827 0.03033 Luminescence 1 1500- Net ear thickness (x 0.01 mm) 1000- 500- 15- 0.1 IL-4 Sham- Reduction in mouse ear inflammation HL.. Vehicle Crisabarole (Img)- Tacrolimus (0.05mg) -Donor 1 Donor 2 PanTrk (0.5mg)- Dexamethasone (0.1mg)- Benevolent 31#32BEN-2293 - Phase Ib safety and tolerability study successfully completed Phase Ib/Ila double-blind, placebo-controlled, first-in-patient, two-part clinical study Phase Ib 2 3 First-in-human dose escalation Phase Ib completed Dec 2021 8 Mild-Moderate AD patients (18-65 years) per cohort, randomised 3:1 BEN-2293: Placebo Safety, Tolerability, PK • Adaptive multiple ascending dose cohort design . Includes efficacy endpoints • MALDI imaging To evaluate human skin PK Phase Ib: ✔ Successfully completed Safety and Tolerability arm BSA= Body surface area. QD = once per day dosing BID= twice per day dosing Cohort 1 0.25% QD 7 Days 10% BSA Not tolerated/ Inadequate SM Adaptive design for Phase lb of the clinical study Tolerated/PK Supports SM Tolerated/PK Supports SM Cohort 2 1% QD 7 Days 10% BSA Cohort 3a 1% QD 14 Days 30% BSA Terminate Study BenevolentAl Proprietary Not tolerated/ Inadequate SM Not tolerated/ Inadequate SM Cohort 3b 0.25% QD 14 Days 30% BSA Not tolerated/ Inadequate SM Terminate Study Tolerated/PK Supports SM Tolerated/PK Supports SM Cohort 4a 1% BID 14 Days 30% BSA Cohort 4b 0.25% BID 14 Days 30% BSA Benevolent 32#33BEN-2293 - indicative data from Phase Ib Caveats: • Phase lb was NOT powered to meaningfully assess efficacy - only 6 patients dosed with active per group • Maximum duration of dosing 14 days (EASI score changes typically measured at 28 days) Mean change from baseline Mean Change from Baseline %BSA affected in treated areas 14 day placebo treatment 1 0 -4 -5 -6 0 Cohort 3 Patients dosed 1% ointment, 30% BSA, once per day 1 2 3 Cohort 4 Patients dosed 1% ointment, 30% BSA, twice per day 4 EASI: Eczema Area and Severity Index 5 6 7 Day 8 9 10 11 12 13 14 15 BenevolentAl Proprietary 0.25% QD 10%BSA 7d 1% QD 10%BSA 7d 1% QD 30%BSA 14d 1% BID 30%BSA 14d PBO 7d PBO 14 d Benevolent 33#34BEN-8744 - Phosphodiesterase 10 (PDE10) - a novel target for UC Transcriptomics data support the rationale for PDE10 as a novel target for UC PDE10 regulates signal transduction by hydrolysing cGMP PDE10 is significantly upregulated in UC-derived colon and colonic mucosa samples, whilst guanylyl cyclase, which makes cGMP, is down-regulated ● Reduced levels of guanylyl cyclase correlate with increased TNF-a in UC colonic mucosa (¹) CGMP is downregulated in UC and its expression inversely correlates to disease severity ● PDE10 is well-studied in CNS disorders but not in inflammation with zero linkage to UC PDE10 demonstrates restricted expression in peripheral tissue Reduces the safety liability of targeted inhibition nitric oxide synthase arginine soluble guanylyl cyclase NOS NO readily diffuses across plasma membranes NO activates GC GTP CGMP activates protein kinases and other proteins breaks down phosphodiesterase e.g. PDE10 PDE10 degrades CGMP Source: (1) Brenna et al. Scand J Gastroenterol 2015 Image (left): Rashed, Second Messenger System 2018 GUCYZC ' • LogFC POE10A LogFC Differential RNA expression of PDE10A and GUCY2C: normal vs UC colonic mucosa colon BenevolentAl Proprietary ܣ ܝ ܗ ܗ ܬ ܝ ܚ ܝ ܘ 0060 + M N TO log FPKM Adipose Tissue Adrenal Gland Bladder Blood Blood Vessel Brain Breast Cervix Uteri Colon Esophagus Fallopian Tube Heart Kidne RNAseq Lume Tissue Nerve M Ovary Pancreas Pituitary Prostate Salivary Gland Skir Small Intestine Spleen Stomach Testis Low basal PDE10 expression, highest levels in brain. Low basal soluble guanylate cyclase (GUCY2C) levels except in colon and sm intestine Gene m PDESA PDE 10A GUCY2C Benevolent 34#35PDE10 inhibition: potential to normalise dysregulated mechanisms in IBD Macrophage M₂ M1 Dendritic Cell INF-KB PGE₂ IL-12 IL-6 IL-18 IL-1B IL-23 TNF Th1 IFNY TNFO | IL-2 ILC3 TIL-17A IL-22 Th17 IL-17A IL-17F IL-21 IL-22 IL-23 • Reduced inflammatory cytokine release from intestinal epithelia via ↓NFkB(¹) • Reduced tissue-resident macrophage activation (1) Reduced intestinal inflammation PDE10 inhibition ↑ CGMP ↑ CAMP Sources: (1) Harmel-Laws et al PLoS One 2013: (2) Han et al PLoS One 2011; (3) Brenna et al Scand J Gastroenterol 2015 Images: Nettleford and Prabhu, Antioxidants 2018 (left); He et al. Int J Mol Sci 2020 (right) Na Nutrients autong mas mara claudins PAMR MLCK fluid Nutrients occludin MLCK TNF • Improved TJ assembly via PKG/PKA-mediated pMLC (2) • Improved fluid/mucus homeostasis via PKG phosphorylation of intestinal CFTR (3) Improved barrier integrity Benevolent 35#36Strategic validation: successful collaboration with AstraZeneca Multi-year Target-ID collaboration is delivering multiple, novel targets for complex diseases with high unmet need Separate data environment established to integrate AstraZeneca's data into a bespoke Knowledge Graph BenevolentAl and AstraZeneca teams working in close collaboration to explore, identify and validate targets Deal structure of upfront license fee, milestone payments and downstream royalties Data generated via the collaboration enriches the Benevolent Platform™ BenevolentAl Proprietary AstraZeneca Five novel targets selected for AstraZeneca's portfolio to date 2019 Initial deal focussed on Chronic Kidney Disease & Idiopathic Pulmonary Fibrosis 2022 3-year collaboration expansion to include Heart Failure & Systemic lupus erythematosus#37Using our platform for wider societal benefit Identified a COVID-19 treatment now approved for use by the FDA RAPID Identified baricitinib as a treatment in just 48 hours, published research in The Lancet in Feb 2020 NOVEL Our technology and Al workflows identified a previously unknown antiviral mechanism (¹) EFFECTIVE COV-BARRIER trial showed baricitinib reduces mortality by 38% in hospitalised patients(2), and by 46% in ventilated or ECMO patients (3) FDA FDA approved the baricitinib to treat COVID-19 in May 2022(4) after first granting EUA in Nov 2020 (5) Lilly Led to equity investment from Eli Lilly BenevolentAl Proprietary DNDi Drugs for Neglected Diseases initiative Non-commercial collaboration → Focused on Dengue fever - a major healthcare burden → Aims to deliver biological targets and drug repurposing candidates → Experimental validation in progress - 6 assays#38Categorisation of AI/ML companies in biotech, hit & target-ID Companies can be characterized across two key dimensions: Original technology focus and drug discovery approach Target-ID WHAT TARGET DO WE NEED TO HIT TO BE EFFECTIVE AND SAFE IN A SPECIFIC DISEASE (pathways, cellular processes)? HIGH COMPLEXITY THROUGH BIOLOGY Many layers of knowledge needed, and many areas where research is not yet complete Hit-ID HOW DO WE NEED TO HIT THE TARGET WE HAVE IDENTIFIED (specific drug characteristics)? HIGH COMPUTATIONAL COMPLEXITY Atom-to-atom interaction is relatively well known, but requires many calculations and simulations Source: Company Websites, Oliver Wyman Analysis Company archetypes Target ID Original technology application focus Hit ID A Benevolent SCHRÖDINGER. Hypothesis driven RECURSION Exscientia RELAY THERAPEUTICS Drug Discovery approach Figure: Oliver Wyman Analysis (listed companies only) BenevolentAl Proprietary B Non-Hypothesis driven C Hypothesis driven Involves a data-driven hypothesis-led approach to therapeutic target identification Non hypothesis driven Leverages technology to identify solutions without specific conditions to target specified at the outset Pharma companies also active in the space, through internal development and/or collaborations Benevolent 38