ENTRAIntelligence
BRIEFINGRTOWALL STREETNYC AISEP 26, 2026
All Briefings

Wall Street AI Teams Face the RTO Premium Calculus

Goldman Sachs, JPMorgan, and Citadel enforce five-day attendance for AI and quant roles, generating a 39% in-office TC premium — but NYC hidden costs absorb most of it.

+39%Finance AI in-office TC premium, NYC · Q3 2026

The 39% in-office total compensation premium for finance AI roles in New York City is the largest ENTRA has recorded since tracking began in Q1 2024. A mid-senior Machine Learning Engineer at Goldman Sachs or JPMorgan Chase earns a median $312,000 in base pay against $224,000 at an equivalent-level role at Cohere's Manhattan office or Scale AI's New York team — and that gap widens further when bonuses and equity are included. But after New York State and City income taxes, the cost of housing near a Midtown or Lower Manhattan office, daily commute, and market-rate childcare, the annual net disposable income advantage for the median in-office finance AI practitioner with two dependents compresses to under $8,000. Wall Street built the best-compensated AI jobs in New York. The city is collecting most of the surplus.

How Goldman, JPMorgan, Citadel, and Two Sigma Define the Mandate

No major Wall Street institution operating AI and quantitative research teams in New York City offers those teams hybrid flexibility. The policy positions, as of September 2026, are uniform and explicit.

Goldman Sachs has required five-day attendance for its Strats division and core Engineering functions since 2023. That policy was not softened during the 2025 hybrid debates that moved Deutsche Bank and Barclays toward two-day minimums. Goldman's Q3 2026 AI postings — Quantitative Researcher (Core AI/ML), NLP Engineer (Market Intelligence), AI Risk Analyst, and Machine Learning Engineer (Trading Infrastructure) — all carry a "New York — In-Office" designation on the firm's career portal. As of September 19, 2026, zero hybrid-eligible postings were active across those four role types, per ENTRA Job Signal Index Q3 2026 monitoring of Goldman's public career pages.

JPMorgan Chase formalized its global five-day mandate on January 10, 2025. The policy applies to its AI Research and Data Science division — approximately 2,100 engineers and researchers in New York City working on LLM-powered risk frameworks, algorithmic trading overlays, and enterprise AI tooling. JPMorgan posted 47 AI and ML roles in New York City in Q3 2026, per ENTRA Job Signal Index monitoring. The postings for Machine Learning Engineer (Trading), Quantitative Researcher (AI Research), and AI Risk Analyst now carry a "Hybrid Flexibility: None" clause in the benefits section — language that appears new to Q2 2026, suggesting the firm moved from obscuring the policy to advertising it.

Citadel operates full five-day attendance at its 52nd Street Manhattan headquarters. Quantitative Researcher and ML Engineer (Systematic Trading) roles sit among the highest-compensated AI positions in the US market — base bands starting at $250,000 for senior researchers, with principals regularly clearing $1.5M–$2.5M total comp. Ken Griffin's stated view, expressed publicly in 2023 and not revised since, is that remote arrangements compromise both employee and firm. As of Q3 2026, no Citadel AI role in New York carries an alternative.

Two Sigma runs the most flexible posture among the four: four days on-site as official policy for Quantitative Researcher and ML Researcher roles at its Hudson Yards office, with Fridays nominally discretionary. ENTRA Q3 2026 Recruiter Survey respondents who moved through Two Sigma's hiring process described an effective five-day norm regardless. For practical purposes, Two Sigma belongs in the same policy bloc as the others.

What the AI-Native Alternatives Actually Pay

Three AI-native employers maintain operating presences in Manhattan large enough to constitute a genuine alternative for finance AI practitioners — and none of them require five days.

Cohere's Midtown Manhattan office runs on a two-day minimum hybrid schedule. The firm posted 11 New York-based AI engineering and applied research roles in Q3 2026. Compensation bands for Cohere's NYC Applied ML Engineer and Research Engineer roles run $290,000–$420,000 base plus $140,000–$200,000 equity on a four-year vest. At the senior end, total comp approaches $620,000 — approximately 57% of the Goldman Strats equivalent.

Scale AI's New York team operates on a project-based presence model: on-site during major client review cycles and annotation campaigns, roughly two to four concentrated weeks per quarter rather than five days every week. Scale's NYC AI Quality Lead and Data Science roles posted in Q3 2026 carried bands of $220,000–$310,000 base with equity participation by level. Scale does not compete with Wall Street on total compensation. It competes on structure — specifically, the ability to work from home for the equivalent of 46 of 52 weeks per year.

Anthropic's New York office remains deliberately contained: fewer than 40 full-time employees as of Q3 2026, concentrated in AI Policy Research, Enterprise Solutions Engineering, and Partnerships functions. For a Goldman NLP Engineer weighing lateral movement without relocating to San Francisco, Anthropic's NYC presence is a live option. It is not a compensation-neutral one. The TC step-down from Goldman Market Intelligence to Anthropic NYC is material.

The Net Calculation That Is Moving Candidates

The 39% gross TC premium is the starting point. What a finance AI practitioner actually pockets after the full New York attendance cost is a different number.

Housing: a one-bedroom apartment within 15 minutes of Goldman's West Street headquarters or JPMorgan's Park Avenue campus rents for $4,800–$5,500 per month in Q3 2026, per StreetEasy median data for Tribeca, Battery Park City, and Murray Hill. The equivalent unit in Long Island City or Hoboken — viable for a hybrid employee with two or three in-office days — runs $3,100–$3,600 per month. Annual delta: $20,400–$28,800.

Transit: a 30-day unlimited MetroCard costs $132. A full five-day commuter with occasional car service supplements for late trading nights spends $3,000–$3,800 annually in transit costs. A hybrid employee traveling two days per week spends approximately $1,400.

Childcare: Manhattan center-based infant and toddler care costs $2,800–$3,800 per month, per the NYC Administration for Children's Services 2026 Cost Report. For a practitioner commuting five days, center-based care is not optional. The annual line item runs $33,600–$45,600. A remote-flexible parent in the same metro using a part-time home arrangement or school-age programs pays closer to $15,000–$20,000 per year.

The tax overlay compresses what remains. New York State's top marginal rate is 10.9%; the New York City resident surcharge adds 3.876%. At a $312,000 base salary — the median in-office finance AI base in this analysis — the combined federal, state, and city marginal burden on the incremental compensation above the AI-native firm equivalent runs approximately 51 cents per dollar. The gross TC advantage for the median mid-senior practitioner is approximately $88,000. After taxes, $43,100. After incremental housing ($20,400), transit ($1,600), and childcare ($13,600) costs attributable directly to five-day attendance, the net disposable income advantage is $7,500 — below the $8,000 threshold for most mid-senior practitioners with dependents. For practitioners without childcare costs, the figure is closer to $25,000 — real money, but not the $88,000 gross gap that recruiter conversations typically anchor to.

What Recruiters Are Hearing in Q3 2026

Fifty-eight percent of NYC finance AI candidates currently in active recruiter pipelines cited RTO policy as the primary reason for initiating a job search, per the ENTRA Q3 2026 Recruiter Survey (n=312, NYC-based finance AI practitioners, collected August–September 2026). That figure was 41% in Q1 2026 — a 17-percentage-point rise over two quarters. Among candidates with children under five, the share citing RTO as the primary trigger reached 71%.

The recruiter-side picture is consistent: finance AI roles in New York are taking longer to fill. ENTRA Q3 2026 data shows median time-to-offer for a Goldman Quantitative Researcher role at 34 days, up from 26 days in Q3 2025. JPMorgan's Machine Learning Engineer (Trading) median time-to-offer is 39 days, up from 28 days. The pool is not shrinking — applications per finance AI posting in NYC rose 14% year-on-year — but the conversion from application to accepted offer is deteriorating. Candidates are entering processes, completing technical screens, and withdrawing at the offer stage. The reason logged most frequently in ENTRA's post-withdrawal recruiter surveys: attendance requirement.

What to Watch

Three signals define the next 90 days in New York finance AI.

Goldman Sachs and JPMorgan's December compensation review cycles will be the first to incorporate post-Labor Day 2026 retention and acceptance-rate data. Both firms typically set 2027 band ranges in November and publish updated compensation frameworks to managers in early December. If Q3 attrition in AI roles has moved materially — and ENTRA's data suggests it has, at Goldman specifically — the December cycle is when finance AI comp will respond. The question is whether band increases or attendance flexibility adjusts first. History at both firms points toward comp.

Cohere's Manhattan headcount trajectory is a leading indicator of whether AI-native firms can absorb the talent that Wall Street mandates are pushing out. Cohere posted 11 NYC roles in Q3 2026. If that figure rises above 20 in Q4, it signals the firm is actively building a New York engineering team large enough to function as a genuine alternative talent sink.

The childcare math will not get easier. New York City's subsidized childcare expansion, funded through the 2025 State Budget childcare package, has a waitlist of approximately 14,000 families as of September 2026. Until subsidized capacity catches up with demand, the five-day attendance requirement at Wall Street firms will continue to land disproportionately on finance AI practitioners with young children — the precise cohort that AI-native firms are designed to recruit.

The 39% premium is not a fiction. It is, for most recipients, a tax on the city itself.


Methodology note: Sources — ENTRA Job Signal Index Q3 2026, monitoring approximately 2,100 active AI and ML postings across New York City's 20 largest financial and technology employers; ENTRA Q3 2026 Recruiter Survey (n=312, NYC-based finance AI practitioners, collected August–September 2026); StreetEasy Q3 2026 Manhattan and Queens rental medians; NYC Administration for Children's Services 2026 Cost Report; Levels.fyi Q3 2026 compensation data for NYC AI engineering roles; New York State Department of Taxation and Finance 2026 rate schedules.

ENTRAGlobal Career Platform

Find AI talent. Find your next role.

Booking is hotels. · Airbnb is apartments. · ENTRA is global careers.

Open ENTRA Careers
End of article

ENTRA Intelligence is independent media on global hiring. Reach the editor at intelligence@entracareers.com

ENTRAGlobal Career Platform

Find AI talent. Find your next role.

Booking is hotels. · Airbnb is apartments. · ENTRA is global careers.

Open ENTRA Careers
More from ENTRA Intelligence

Continue reading.

BRIEFINGEurope · AI
€92KMedian annual total compensation for senior AI evaluator (German-language specialist) · Berlin · Q3 2026

Berlin's AI Evaluation Premium: How Works Council Law and GDPR Built Germany's Trainer Economy

German Works Council legislation — designed to protect workers, not build AI infrastructure — has become an inadvertent engine of AI trainer professionalization. Companies cannot deploy annotation contractors at scale without Betriebsrat consent, forcing structured employment that commands €75,000–€110,000 annually in Berlin's evaluation market.

OCT 8, 2026berlingermany
BRIEFINGMiddle East · AI
3.4×Arabic RLHF specialist TC premium over equivalent English-only evaluator · UAE · Q3 2026

Dubai's Arabic RLHF Premium: How the UAE Became the Gulf's Indispensable AI Training Hub

Arabic-English bilingual AI evaluators in Dubai command AED 55,000–85,000 per month — tax-free, comparable to $178K–$276K annualized — driven by scarcity of Arabic linguists with the ML literacy frontier labs need. Dubai's dialectal diversity, not Riyadh's infrastructure, is the Gulf's structural edge.

OCT 8, 2026dubaiarabic-rlhf
ANALYSISUSA · IT
$350KMedian total comp for senior platform engineer at leading AI training infrastructure company · US · Q3 2026

The Platform Engineering Stack Behind the AI Trainer Economy

RLHF at frontier scale is a distributed systems problem masquerading as a labor economics story. The platforms building the AI trainer economy — Scale AI, Mercor, Labelbox — have quietly become some of the most technically demanding annotation infrastructure in the industry, with engineering teams commanding $280K–$420K to solve consensus algorithms for human preference data at 100,000 annotations per day.

OCT 8, 2026platform-engineeringrlhf-infrastructure