|
Listen to article
|
OpenAI’s robotics chief quit Saturday over the company’s agreement to provide artificial intelligence systems to the Pentagon’s classified network, becoming the most senior departure yet from a deal that has divided Silicon Valley over the boundaries of military AI deployment.
Caitlin Kalinowski said surveillance of Americans without judicial oversight and autonomous weapons without human authorization crossed lines that deserved more deliberation than they received.
She joined the company in November 2024 after leading augmented reality development at Meta and had been overseeing technical work in robotics.
OpenAI defended the Defense Department arrangement, saying it creates a workable path for responsible national security uses while establishing clear prohibitions against domestic surveillance and autonomous weapons. The company acknowledged that employees and outside groups hold strong views on the issues and said it would continue discussions with stakeholders worldwide.
Read also: Carney Urges Australia-Canada Unity As Global Order Weakens
The February deal followed collapsed negotiations between the Trump administration and Anthropic, which pressed for assurances its technology would not be used for mass surveillance or fully autonomous weapons.
When those talks broke down, President Donald Trump ordered federal agencies to stop working with Anthropic and the Pentagon declared the company a supply-chain risk.
That designation, previously reserved for adversaries like China’s Huawei Technologies, prompted Anthropic to announce it would challenge the classification in court. OpenAI publicly disagreed with the blacklisting.
Chief Executive Sam Altman said earlier this month that his company’s rush to finalize the Pentagon contract looked “opportunistic and sloppy.” The admission reflected internal tensions over whether OpenAI moved too quickly to capitalize on Anthropic’s exclusion from government work.
Read more: Google Sued After Gemini AI Instructed Man To Kill Himself
Kalinowski’s resignation underscores divisions within the AI industry over military applications. Some researchers argue that keeping advanced systems out of defense hands creates risks if adversaries develop similar capabilities without ethical constraints.
Others maintain that participating in weapons development normalizes uses that could prove catastrophic.
OpenAI has not said whether its Pentagon services will replace work previously performed by Anthropic or supplement existing capabilities. The company also has not disclosed which specific AI models will be deployed on classified networks or what oversight mechanisms will govern their use.
TechCrunch first reported Kalinowski’s departure. She wrote on X that artificial intelligence has an important role in national security but that certain applications require more debate than they received before contracts were signed.
The controversy arrives as Congress considers legislation that would establish guardrails for military AI while encouraging development to maintain technological advantages over China and Russia. Lawmakers have struggled to balance innovation incentives with safeguards against autonomous weapons that could select and engage targets without human approval.
OpenAI’s position as the industry’s highest-profile company makes its Pentagon relationship symbolically significant beyond the contract’s immediate scope.
Competitors and researchers will watch how the arrangement develops and whether promised restrictions hold once systems are deployed in operational environments.
