OpenAI Robotics Chief Caitlin Kalinowski Resigns Over Rushed Pentagon AI Deal

In a major blow to OpenAI’s hardware ambitions, Caitlin Kalinowski, the company’s Head of Robotics and Consumer Hardware, has resigned. Her departure is a direct protest against OpenAI’s recently signed agreement with the U.S. Department of Defense (Pentagon), a deal she claims was fast-tracked without essential ethical safeguards or governance oversight.

Kalinowski, a high-profile hire from Meta who joined in November 2024, is the most senior executive to exit following the company’s pivot toward military partnerships.

1. The Resignation: "Principle, Not People"
Kalinowski announced her resignation via social media on Saturday, March 7, framing her exit as a matter of fundamental ethics.

The "Red Lines": She specifically cited concerns over lethal autonomy (AI making life-or-death decisions without human intervention) and domestic surveillance of Americans without judicial oversight.

Governance Failure: In a follow-up post, she emphasized that her primary issue was procedural. "My issue is that the announcement was rushed without the guardrails defined. It’s a governance concern first and foremost. These are too important for deals or announcements to be rushed."

Personal Stance: Despite her exit, she expressed "deep respect" for CEO Sam Altman, stating the move was about "principle, not people."

2. Context: The Pentagon Deal and the "Anthropic Gap"
The resignation highlights a growing divide in Silicon Valley over military AI. OpenAI’s contract was reportedly secured under controversial circumstances:

The Anthropic Rejection: In late February, rival firm Anthropic refused the Pentagon’s terms, citing a lack of bans on autonomous weapons and domestic surveillance.

The "Supply-Chain Risk" Label: Following that refusal, the Pentagon designated Anthropic a "supply-chain risk." OpenAI moved quickly to fill the void, reaching an agreement just hours later to deploy its models on the military’s classified cloud networks.

Scope of Use: The deal officially allows the technology to be used for cybersecurity, logistics, and intelligence analysis, but internal critics fear these boundaries are too porous.

3. Internal and External Fallout
The Pentagon deal has triggered a cascade of negative reactions both inside and outside the company:

Executive Exodus: Hours after the deal was confirmed, Max Schwarzer, OpenAI’s VP of Research (who led the development of GPT-5), also resigned to join Anthropic.

Subscriber Backlash: Reports suggest approximately 1.5 million users canceled their ChatGPT Plus subscriptions in the week following the announcement, citing privacy and ethical concerns.

Employee Dissent: Over 330 employees from OpenAI and Google signed an open letter supporting Anthropic’s refusal to accept unconditional military terms.

4. OpenAI's Official Defense
OpenAI has moved to damage-control mode, with Sam Altman admitting that the deal’s rollout was "opportunistic and sloppy."

Self-Imposed Restrictions: The company maintains that its "red lines" are still in effect, legally prohibiting the use of its models for mass domestic surveillance or the development of fully autonomous weapons.

Technical Safeguards: OpenAI claims its "cloud-only" deployment allows it to retain a "safety stack" that can shut down the model if it detects a violation of use-case policies by the military.

5. Impact on OpenAI Robotics
Kalinowski’s departure is a significant setback for OpenAI’s physical AI program.

Household Robotics: She was leading a high-profile project in San Francisco training robotic arms for household tasks.

Ive Collaboration: She was also a key link in the collaboration with former Apple designer Jony Ive to develop a dedicated AI consumer hardware device, scheduled for late 2026. Her exit leaves the hardware division without its most experienced veteran.

Industry Quote: "This resignation signals that the 'safety-first' culture OpenAI once championed is in a state of civil war. When your head of robotics says the guardrails aren't there, the market has to ask: who is actually in control of the machine?" — Analyst, Gartner Research