AI Hiring Brief · 2026-07-10
19 frontier AI labs · 99 roles added in 24h · agihunt.info/en/jobs
Today at a glance
- Anthropic goes deep into biology: Four life sciences roles posted in a single day — "Research Associate, Biology," "Research Scientist, Life Sciences (Experimental Biology)," "Research Operations Lead, Biology," and "Partnership Manager, Life Sciences" — spanning bench science, operations, and commercial partnerships simultaneously. No single day in recent memory has seen this level of concentrated biological hiring from Anthropic.
- Apple's tactile robotics cluster: Three simultaneous postings in the Special Projects (Robotics) group in Santa Clara — "Mechatronics Research Engineer," "Senior Tactile Research Engineer," and "Senior Tactile Embedded Controls Research Engineer" — signal that Apple's robotics program has moved beyond concept into precision manipulation hardware.
- OpenAI builds commercial geography and physical infrastructure in parallel: Vertical sales roles for insurance and banking, regional leads for India and Japan, an EU policy post in Brussels, and an APAC controller in Singapore run alongside four Scaling department hires covering civil, project, mechanical, and electrical engineering for data centers.
- Google DeepMind threads biosecurity and tool-use research: "Threat Modeler Lead, CBRN" landed the same day as "Research Scientist, Protein Design" and "Research Scientist, Gemini Tool Use" — simultaneous bets on capability building and misuse containment.
- NVIDIA bets on silicon photonics: A "Senior Silicon Photonics Design Engineer" posting points to optical interconnect investment as a response to bandwidth bottlenecks emerging at next-generation AI cluster scale.
Company moves
OpenAI added 22 roles, the largest single-day count among global labs. The Go To Market cluster spans "Enterprise Sales Manager, Insurance" and "Enterprise Sales Manager, Banking" (both New York), "Manager, Solutions Engineering - Digital Natives" (Dublin), "Account Director, Startups" (Sydney), "APAC Controller" (Singapore), "Product Partnerships Lead, India," and "Enterprise Marketing Lead, Japan" — a geography-by-geography build-out of the commercial layer. Policy coverage is simultaneous: "EU Policy & Partnerships Manager" (Brussels) and "U.S. State Policy & Partnerships Manager, Global Affairs" address both supranational and subnational regulatory relations. The Scaling department added four physical-infrastructure engineers — "Civil Engineer," "Project Engineer," "Data Center Research & Development Engineer, Mechanical - Industrial Compute," and "Data Center Design Engineer, Electrical - Industrial Compute" — roles that belong to permanent facility design and oversight, not software-defined infrastructure.
Anthropic posted 10 roles, four of them forming a self-contained biology team: bench science ("Research Scientist, Life Sciences (Experimental Biology)"), logistics ("Research Operations Lead, Biology"), recruiting pipeline ("Research Associate, Biology"), and commercial interface ("Partnership Manager, Life Sciences"). On the engineering side, "Engineering Manager, Agent Runtime Platform" (Infrastructure) and "Staff Engineer, Datacenter Server Lifecycle" (Sydney) continue the build-out of the agent execution layer and global data center footprint.
Google DeepMind added 10 roles. "Research Scientist, Protein Design" (London) and "Threat Modeler Lead, CBRN" (New York/Mountain View) are the headline pair. "Research Scientist, Gemini Tool Use" makes tool-calling a standalone research specialization. "UX Research Operations Lead, GeminiApp" rounds out the consumer-facing operational layer.
Microsoft AI led all companies with 27 additions against 32 closures — active rotation in a large apparatus. Notable new postings include "Principal Applied Scientist (Multiple Openings) - Copilot and Agents Core" (Redmond) and "Senior CAD Engineer" in Hardware Engineering. Japan received two hires: "Senior Manager Data Center GTM" and "AI Business Solution - AI ビジネスプロセスシニアソリューションスペシャリスト (大手エンタープライズ企業担当)" (senior AI business process solution specialist for large enterprise accounts) — localized headcount at the enterprise deal level.
NVIDIA posted 7 roles. Beyond silicon photonics, the batch includes "Senior DFT Engineer" (design-for-testability, a chip complexity indicator), "Customer Program Manager - Automotive" (dedicated automotive AI customer management), and "Solutions Architect, Model Builder - LATAM" (Latin America go-to-market).
Apple AIML added 5 roles, three robotics-focused in Santa Clara. The remaining two — "ML Engineering Program Manager, Apple Ads" and "Machine Learning Engineer, Proactive" (Cupertino) — indicate continued on-device ML investment for ads relevance and proactive assistant features on a separate track.
Baidu AI (百度 AI) posted 11 new roles against 40 closures, a net of −29 — the steepest single-day drawdown in today's dataset. New technical hires include "大模型评测算法工程师" (large model evaluation algorithm engineer) and "大模型算法工程师" (large model algorithm engineer, applied model R&D). Commercial additions include enterprise and cloud AI sales roles across Beijing and Hebei.
ModelBest (面壁智能) added 5 roles, all in market brand: "品牌视觉专家" (brand visual expert), "品牌传播经理(行业PR方向)" (brand communications manager, industry PR), "品牌传播经理(财经PR方向)" (brand communications manager, financial PR), "媒体关系经理" (media relations manager), and "品牌策略专家" (brand strategy expert). Five brand hires with zero engineering additions is the profile of an organization preparing for a public-facing announcement.
StepFun (阶跃星辰) added one role: "测试开发工程师(APP方向)" (app-facing test development engineer). Meta AI added one: "Computer Vision Engineer" (Burlingame, CA). Both are routine fills.
SpaceXAI, Perplexity, Mistral AI, Cursor, Lovable, Moonshot AI (月之暗面 Kimi), DeepSeek (深度求索), Zhipu AI (智谱 AI), and Meituan LongCat (美团·龙猫) posted no additions. Meituan LongCat closed 11 roles against zero additions — the largest net drawdown among Chinese labs today.
What it signals
The convergence of Anthropic and Google DeepMind on biology within the same 24-hour window carries more weight than coincidence. Anthropic is not merely hiring computational biologists — "Research Scientist, Life Sciences (Experimental Biology)" is a wet-lab hire, capable of generating the ground-truth data that model training requires. DeepMind's "Research Scientist, Protein Design" follows a similar logic. Both labs have made AI-for-biology claims publicly; they are now staffing to deliver internally rather than through external partnerships. Anthropic's 7-day figure of 424 additions suggests this is a sustained intake, not a one-day surge.
Google DeepMind's same-day "Threat Modeler Lead, CBRN" alongside its biology and protein research hires is the most pointed safety signal in today's data. CBRN (Chemical, Biological, Radiological, Nuclear) threat modeling is the containment-side correlate of biological AI capability — as labs accumulate the capacity to accelerate biology research, the requirement to model its misuse grows proportionally. The parallel appearance of capability-building and containment hires at the same lab on the same day is a structural tell: the frontier is no longer treating safety as a downstream activity.
OpenAI's Scaling department additions — civil and project engineers alongside data center mechanical and electrical engineers — confirm that its compute infrastructure ambitions have outgrown third-party delivery. These are domain-specific permanent hires for physical facility design and oversight. With 771 additions over 7 days and 716 open roles, OpenAI's breadth across commercial, policy, engineering, and infrastructure remains unmatched among the 19 companies tracked here.
Among Chinese labs, the divergence is notable. Baidu AI's −29 net today continues a pattern of active role rotation rather than net expansion. ModelBest's brand-only surge stands apart: five communications and PR hires with no technical additions is not a growth signal, it is a readiness signal — the organization believes it has something to announce soon. The remaining Chinese lab cohort largely sat out today, consistent with a quieter hiring cadence relative to global peers, though their 7-day open counts remain substantial across the board.