GSK's 2025 Gender Pay Gap Report, published under mandatory UK reporting with a mean gap of 18.3% across its full UK workforce, is the most detailed compensation disclosure the Stevenage AI engineering function has produced. It does not contain a single ML-specific salary band. Across the Oxford-Cambridge Arc, which clusters approximately 180 ML researchers at AstraZeneca's global headquarters at Granta Park, Cambridge; 140 engineers at GSK's AI lab in Stevenage; roughly 300 staff at BenevolentAI in London; and 350 at Exscientia on Oxford Science Park, that combination of organisational gap data without role-level band disclosure describes the dominant posture of UK life sciences AI employers in August 2026. AstraZeneca is the exception: ENTRA Q2 2026 monitoring found that 72% of AZ's UK AI roles carried a salary range at posting, the highest disclosure rate among named life sciences employers on this corridor. The gap between the best and the norm is not primarily legislative. No UK law mandates salary bands in job advertisements before 2027 at the earliest. It is a competitive choice made in a market where life sciences AI employers already cede 15-20% on base salary to London finance and pure-tech AI firms operating a short train ride away at King's Cross.
The Corridor's Compensation Baseline
ML researchers at pharmaceutical AI functions along the Oxford-Cambridge corridor earn between £75,000 and £120,000 base depending on seniority and specialisation, per ENTRA Q2 2026 monitoring. The comparison point is unambiguous: a Senior ML Engineer at a King's Cross AI corridor employer commands £90,000 to £145,000 base plus equity acceleration. Life sciences AI runs 15-20% below that at mid-seniority levels. Both cohorts clear the Skilled Worker visa salary floor of £38,700 by multiples; visa economics are not the constraint. The constraint is a base salary gap that becomes precise and visible the moment life sciences AI employers post salary bands, which is precisely why some of them have not.
The benefits profile diverges structurally from pure-tech. Pharma employers on the Arc offer defined-contribution pension schemes above the 5% auto-enrolment floor, typically 8-12% employer contribution, and share schemes structured through Save As You Earn or Company Share Option Plans rather than the cliff-vest RSU model common at London AI startups. For candidates who intend to remain in the UK for five or more years, the total compensation comparison shifts materially relative to a raw base comparison. The argument is legitimate. It is also easier to make when the base number is disclosed.
AstraZeneca: The 72% Disclosure Standard
AstraZeneca's AI platform function at Granta Park, Cambridge, covers approximately 180 researchers working across drug discovery, generative chemistry, and clinical trial AI. The function has recruited heavily from the Cambridge ML PhD pipeline and from Sanger Institute bioinformatics alumni. Per ENTRA Q2 2026 postings analysis, AZ posts salary ranges on 72% of its UK AI roles, a rate that leads the life sciences sector and reflects in part a de facto US-format disclosure norm propagating through AZ's global HR systems. Its disclosed Senior ML Scientist band of £90,000 to £125,000 (~$114K-$158K at Q2 2026 rates) gives candidates enough data to benchmark against London tech alternatives before first contact. The 28% of AZ AI roles that carry no salary range are concentrated at Director level and above, where total compensation including long-term incentive plans diverges substantially from base-only disclosure.
GSK Stevenage: Organisational Data, No ML Bands
GSK's AI and ML engineering function, operating from its Stevenage headquarters with approximately 140 engineers across drug discovery AI, clinical modelling, and digital health, published its 2025 Gender Pay Gap Report in April 2026 under the Equality Act 2010 (Gender Pay Gap Information) Regulations 2017. The report's 18.3% mean gap is consistent with sector-wide patterns: technology functions within large UK employers routinely show double-digit mean gaps driven by seniority skew, the proportion of senior engineering positions held by men. The AI-specific gap is not separately disclosed; GSK's statutory obligation does not require subdivision by function. Salary bands appear in approximately 40% of GSK's AI-tagged UK roles per ENTRA Q2 2026 monitoring, a rate that lags AstraZeneca by 32 percentage points and that gives candidates less pre-application data at a company competing for the same Cambridge and Imperial ML PhD cohort.
BenevolentAI: Startup Disclosure in Drug Discovery
BenevolentAI, which raised $292 million in total funding and operates from London with approximately 300 employees, publishes salary ranges on 65% of its UK postings per ENTRA Q2 2026 tracking. Its Senior NLP Researcher band of £95,000 to £130,000 (~$120K-$165K) sits at the top of the life sciences AI corridor range and approaches the London tech senior IC band, reflecting BenevolentAI's structural position as an AI-native company recruiting against both pharma labs and frontier AI employers. The company draws from UCL and Imperial ML research alumni in addition to Cambridge, and holds an active Skilled Worker sponsor licence. Pre-IPO equity forms a material part of the total compensation argument at Series B stage; the degree to which that equity structure is disclosed alongside salary bands varies by role seniority.
Exscientia and the NASDAQ Transparency Pressure
Exscientia, the Oxford-based AI drug design company listed on NASDAQ (EXAI) and operating from Oxford Science Park with approximately 350 UK employees, faces transparency pressure from two directions. As a NASDAQ issuer, it files US-format compensation disclosures with the SEC; those filings contain named executive officer pay but do not require engineering salary band disclosure. UK Companies House subsidiary accounts provide statutory figures without role-level detail. The practical effect is that Exscientia's transparency is more granular at the executive level than at the ML researcher level. ENTRA Q2 2026 postings analysis tracked ML Scientist roles at the Oxford campus at £80,000 to £115,000 (~$101K-$146K), but salary bands appeared in fewer than 50% of active postings, a disclosure rate below both AstraZeneca and BenevolentAI within the same corridor.
EU AI Act Compliance: New Pay Band
The EU AI Act's Annex III classification of medical device AI as high-risk has created a technical compliance function that did not exist in structured form at most pharma AI employers before 2025. AstraZeneca, GSK, and Exscientia all develop AI systems with EU-market medical device exposure; Annex III obligations apply to those systems regardless of where the employing entity is headquartered. The resulting roles, ML Compliance Engineer, AI Quality Assurance Lead, Regulatory AI Scientist, are appearing at pay bands of £85,000 to £130,000 across the corridor per ENTRA Q2 2026 monitoring. These are net-new positions drawing candidates from both AI engineering and pharmaceutical regulatory affairs talent pools. AstraZeneca and BenevolentAI have posted bands for these roles; GSK and Exscientia have not in the majority of postings tracked.
Where the Corridor Loses Candidates
The talent competition runs on two vectors that the corridor's mission argument does not fully neutralise. London finance, senior ML quant roles at Jane Street, Citadel, and the Goldman Sachs London AI platform, pays £130,000 to £220,000 at senior level. US pharma active in UK remote hiring, including Pfizer's global AI platform and Johnson and Johnson MedTech AI, posts senior roles at $150,000 to $250,000. UK-based candidates in active searches benchmark against both.
The corridor's retention mechanism is patient-outcomes mission combined with pre-IPO equity upside. Candidate surveys consistently show that PhD-level hires leaving academia weight mission heavily; the argument is legitimate and non-trivial. It is most effective at the pre-offer stage when the base salary gap is framed inside a total compensation picture. That framing requires the base to be disclosed. The companies on the Arc with the highest voluntary disclosure rates, AstraZeneca at 72% and BenevolentAI at 65%, are the companies best positioned to make the full-picture case before candidates have already accepted an offer elsewhere.
What the Employment Rights Act 2025 Changes
The Employment Rights Act 2025, which received Royal Assent in December 2025, introduces Equality Action Plans as a mandatory obligation from spring 2027 for employers with 250 or more staff. AstraZeneca UK, GSK UK, BenevolentAI, and Exscientia all cross that threshold. The Plans require documented strategy on gender pay gap reduction, including within technology and research functions, but they do not mandate salary band disclosure in job advertisements. A separate government call for evidence on mandatory salary disclosure in vacancy notices closed on 30 June 2025; legislation is expected before the end of the current parliamentary term per Lewis Silkin LLP analysis published July 2026.
Until that legislation passes, voluntary disclosure is the only mechanism producing band data that ML candidates on the corridor are treating as a pre-application minimum. AstraZeneca's 72% and BenevolentAI's 65% are the current Arc benchmarks. The operational question for GSK and Exscientia is whether waiting for a legal mandate, rather than matching the voluntary standard, produces measurable drag on application quality during the 18-month window before the obligation arrives. The ENTRA Q2 2026 Candidate Survey (n=1,240 UK-based AI job seekers) found that 67% report filtering out listings with no salary range as a routine first step. On the Oxford-Cambridge Arc, that filter is active now.
GSK 2025 Gender Pay Gap Report (18.3% mean gap, full UK workforce) filed April 2026 under the Equality Act 2010 (Gender Pay Gap Information) Regulations 2017, available via the UK Government Gender Pay Gap Service. AstraZeneca salary disclosure rate (72% of UK AI roles carrying a posted salary range) and AZ Senior ML Scientist band (£90,000-£125,000) per ENTRA Q2 2026 postings analysis (n=87 AZ UK AI-tagged roles, April-June 2026). Exscientia ML Scientist band (£80,000-£115,000), BenevolentAI Senior NLP Researcher band (£95,000-£130,000), BenevolentAI disclosure rate (65%), Exscientia disclosure rate (below 50%), and GSK disclosure rate (approximately 40%) per ENTRA Q2 2026 postings analysis (n=340 UK AI and ML-tagged roles across named employers). Life sciences AI ML researcher base range (£75,000-£120,000) and London pure-tech senior IC range (£90,000-£145,000) per ENTRA Q2 2026 UK Salary Survey. London finance senior ML range (£130,000-£220,000) per ENTRA Q2 2026 London Finance AI Salary Monitor. US pharma remote senior range ($150,000-$250,000) per publicly posted role data for Pfizer Global AI Platform and Johnson and Johnson MedTech AI, January-June 2026. EU AI Act Annex III compliance role pay bands (£85,000-£130,000) per ENTRA Q2 2026 monitoring of newly posted regulatory AI roles across 12 UK pharma employers. Headcount figures (AZ approximately 180, GSK approximately 140, BenevolentAI approximately 300, Exscientia approximately 350) per ENTRA Q2 2026 company tracking and LinkedIn headcount signals. BenevolentAI total funding ($292 million across multiple rounds) per public company disclosure. Exscientia NASDAQ listing (ticker: EXAI) per SEC filings. Candidate filter rate (67%) per ENTRA Q2 2026 Candidate Survey (n=1,240 UK-based AI job seekers, April-June 2026, online, self-reported). Employment Rights Act 2025 Royal Assent (December 18, 2025) and Equality Action Plan obligation timeline per Lewis Silkin LLP analysis, July 2026. Skilled Worker visa salary floor (£38,700) per Home Office Immigration Rules, January 2024 uprating, in force August 2026. GBP/USD conversions at approximately £1 = $1.27 per Q2 2026 market data, consistent with ENTRA regional reporting. All salary figures are ENTRA estimates derived from public postings analysis and have not been confirmed by the named employers.
Global Career PlatformFind AI talent. Find your next role.
Booking is hotels. · Airbnb is apartments. · ENTRA is global careers.