OpenAI ended 2025 with roughly 4,000 employees (per The Information), a new corporate structure, and a compensation instrument — the Profit Participation Unit — that had been quietly dying for two years. By March 2026 the company had crossed $25 billion in annualized revenue, closed a $40 billion round led by SoftBank at a $300 billion post-money valuation (later surpassed by a February 2026 round initially reported at $110 billion and subsequently expanded to $122 billion, setting a post-money valuation of $852 billion), and converted every PPU on the cap table to a standard restricted stock unit. The structural conditions that have shaped OpenAI's talent market since 2022 are now materially different, and the hiring push that followed — a plan to nearly double headcount to 8,000 by December — is playing out across three distinct fronts.
What Changed When the PBC Conversion Landed
On October 28, 2025, OpenAI's for-profit subsidiary completed its conversion to OpenAI Group PBC, a public benefit corporation. The capped-profit structure that had governed the company since 2019 — and had made PPUs the defining, and frequently misunderstood, equity instrument of AI recruiting — was dissolved. In January 2026, OpenAI converted existing PPUs to standard RSUs, allocating approximately 25 percent of the new entity's equity to employees and former employees, with the OpenAI Foundation retaining a separate 26 percent stake.
The practical talent implications are significant and underappreciated. PPUs operated under a profit cap that created theoretical upside but genuine complexity in candidate conversations: a prospective hire had to understand what "100x of invested capital" meant, how the waterfall worked, and why the instrument was not a stock option. Senior recruiters at competing labs — particularly Anthropic, which issues standard RSUs — had been using PPU complexity as a consistent offer-stage lever through 2023 and 2024. The RSU conversion eliminates that lever entirely. OpenAI now competes on the same equity language as every other frontier lab.
The conversion also resolved a secondary friction: liquidity sequencing. PPU holders had to wait for a tender offer to realize value. RSU holders at a company with a $300B-to-$852B valuation trajectory — and an IPO that CFO Sarah Friar has flagged for 2027 while Sam Altman has pushed for Q4 2026 — are looking at a fundamentally cleaner path to liquidity. The shift has been reflected in offer acceptance rates: the RSU conversion removed the most-cited offer-stage objection: a candidate no longer needs to model PPU waterfalls to evaluate OpenAI equity against Anthropic RSUs.
One additional compensation mechanism worth noting: in August 2025 — before the conversion closed — OpenAI extended retention bonuses ranging from $300,000 to $1.5 million per person depending on seniority, vesting over two years, to roughly 1,000 research and engineering staff. Average stock grants for the broader employee base reached $1.5 million in 2025. The retention push and the structural conversion were coordinated: the company was simultaneously paying people to stay and removing the most cited reason to leave.
The Revenue Context Behind the Hiring Budget
The headcount expansion is funded by a revenue trajectory that would have been considered implausible eighteen months ago. OpenAI crossed $10 billion annualized in mid-2025, $20 billion by the end of the year, and reached approximately $2 billion per month — $25 billion annualized — by March 2026. Enterprise accounts for roughly 40 percent of that figure and is on track to reach parity with consumer by year-end.
The revenue-per-employee math is compelling for a company in aggressive headcount growth mode. At $3 million-plus in annualized revenue per employee as of early 2026, OpenAI can fund frontier-lab comp bands — median total comp of approximately $608,000, with senior IC bands above $850,000 at the floor and past $2 million at the top — without the financial strain that has constrained smaller labs. The $40 billion round provides additional insurance. This is a company that can spend on talent and has the revenue growth rate to justify doing so.
Three Hiring Fronts That Define H1
Front one: Enterprise deployment. The largest net-new hiring category in H1 2026 is not research — it is customer-facing deployment. OpenAI announced plans to hire hundreds of what the company initially called "Technical Ambassadors" and later rebranded as Forward Deployed Engineers (FDEs): specialists who embed directly into enterprise client organizations to convert pilots into production usage. The role sits at the intersection of solutions engineering, customer success, and applied research. Compensation at the mid-level runs $350,000 to $450,000 total comp; senior FDEs are clearing $450,000 to $550,000, with equity making up 55 to 70 percent of the package at the top of the band. The strategic logic is explicit: OpenAI's own CFO has said that closing the gap between what AI can do and what customers are actually extracting from it is the company's primary focus in 2026. The FDE hire is the operational expression of that bet.
Front two: Safety and preparedness. In December 2025, OpenAI posted a Head of Preparedness role at a base salary of up to $555,000 — a number that was notable enough to generate press coverage, and that reflected the company's intent to compete at frontier-lab comp levels for safety talent, not just research talent. The role was filled in February 2026 by Dylan Scandinaro, who joined from Anthropic's safety team. A Bloomberg opinion piece from March 2026 noted that safety team headcount across AI labs "might fit on a single transatlantic flight" — OpenAI's response has been to hire visibly and expensively rather than defensively. The Preparedness, Safety Systems, and Governance teams are active hiring clusters, and the comp signals suggest OpenAI is treating safety as a talent market where it intends to win rather than merely participate.
Front three: Robotics. On May 31, 2026, Sam Altman announced OpenAI Robotics — a dedicated hardware and physical-AI division — with an immediate public hiring call for "exceptional full-stack hardware, ops, systems, and ML engineers." The careers page currently lists 11 robotics roles in San Francisco, including Actuator Design Engineer, Simulation Applications Engineer, and ML roles tied to distributed data systems. The launch marks OpenAI's most aggressive robotics push since shutting down its original robotics team in 2021. The near-term target is robots for construction and data-center build-out — the infrastructure the AI boom requires. The long-term ambition Altman described publicly is a personal robot for everyone. The hiring profile — mechatronics, control systems, embedded software alongside frontier ML — is a new talent category for OpenAI and suggests competition with hardware-robotics companies such as Agility Robotics, Figure, and Boston Dynamics rather than Anthropic or Google DeepMind.
Three Things to Watch Through H2
The IPO readiness signal. The internal disagreement between Altman (Q4 2026) and Friar (2027) on IPO timing is a talent market variable, not just a capital markets one. A public-company filing creates mandatory equity liquidity for RSU holders. Senior hires evaluating OpenAI offers in Q3 and Q4 will be reading every IPO signal carefully — and competing labs will be using any IPO delay as a counter-offer narrative.
Safety headcount credibility. The Bloomberg piece that characterized AI safety teams as fitting on one plane was aimed broadly, but OpenAI is the company with the most brand-level exposure to safety scrutiny. Whether the Scandinaro hire and the preparedness team expansion translate into a visible headcount commitment — one that can be cited in recruiting conversations — will determine whether OpenAI can recruit safety researchers who might otherwise choose Anthropic on mission grounds.
Altman's "do more with fewer" tension. In January 2026, Altman told staff internally that the company wanted to slow hiring and extract more productivity from existing headcount — a signal that briefly ran counter to the 8,000 target. The tension between a board-level directive to reach 8,000 and a CEO-level instinct toward efficiency will shape how the back half of the hiring plan executes. The FDE and robotics categories are most exposed: both require large net-new teams, and both are in domains where OpenAI has limited institutional hiring muscle.
The PBC conversion removed one constraint. The revenue growth funded the budget. The RSU conversion cleared the equity-language barrier. What H2 2026 will test is whether OpenAI can execute three simultaneous hiring buildouts — enterprise deployment, safety, and hardware — without the recruiter infrastructure and institutional pattern-recognition that Anthropic has been building deliberately for two years.
Sources: OpenAI PBC Conversion — OpenAI — SoftBank Fully Funds $40B OpenAI Investment — CNBC — OpenAI Expects to Increase Headcount to Roughly 8,000 — The Information — OpenAI Plans to Double Workforce — Engadget — OpenAI Achieves Over $20B Annual Revenue Run Rate — MLQ.ai — OpenAI Compensation 2026: PPU Equity Guide — Levels.fyi — OpenAI Salary 2026 — JobsByCulture — OpenAI Research Scientist Salary — Levels.fyi — OpenAI Is Hiring Hundreds of AI Consultants — The Information — OpenAI Plans to Expand to 8,000, Technical Ambassadors — AIBase — Forward Deployed Engineer Compensation Report 2026 — Perspective AI — OpenAI Fills Safety Role at $555K with Anthropic Hire — Bloomberg — AI Companies Talk Safety; Headcount Tells a Different Story — Bloomberg — OpenAI Robotics Hiring — TechFundingNews — OpenAI CFO Sarah Friar on 2026 Practical Adoption — CNBC — OpenAI CFO IPO Timeline Concerns — MLQ.ai — OpenAI $1.5M Retention Bonuses — Calcalist — OpenAI Compensation Overhaul — AInvest