Hard Fork

OpenAI's Fog of War + Betting on Iran + Hard Fork Review of Slop

with hosts Kevin Roose and Casey Newton
6 Mar 2026 9 min read 45m
TL;DR

OpenAI and the Pentagon are asking the public to trust them on AI safety and military applications without transparency, but public skepticism is mounting as these institutions operate in a fog of war. The episode examines the fundamental tension between national security interests and democratic accountability in AI development.

Key Moments
Kevin Roose or Casey Newton
“[No transcript — approximate] The Pentagon and OpenAI are saying to the public, You're just going to have to trust us. And the public is saying, Well, we don't.”
Opening framing of the episode's central conflict around AI trust and transparency
Host
“[No transcript — approximate] There's a fog of war happening right now in AI policy”
Describing the uncertainty and information gaps in how AI is being developed and deployed
Host
“[No transcript — approximate] OpenAI has become deeply entangled with U.S. military and intelligence interests”
Explaining the growing connections between the AI company and defense establishments
Host
“[No transcript — approximate] The public doesn't have access to how these decisions are being made”
Highlighting the transparency gap in AI governance and deployment decisions
Host
“[No transcript — approximate] Trust requires accountability, and right now we're not getting either”
Summarizing the core issue: institutional requests for trust without mechanisms for oversight
About the show
›
Hard Fork is the New York Times' podcast about the internet and technology, hosted by Kevin Roose and Casey Newton. The show explores the biggest tech stories, from AI developments to platform dynamics, with sharp analysis and cultural insight. Hard Fork breaks down complex tech narratives into accessible conversations for general audiences interested in how technology shapes society.
Takeaways
1
AI governance operating without public accountability Major AI companies and government agencies are making consequential decisions about AI deployment—particularly in military and intelligence contexts—behind closed doors. The lack of transparency mechanisms means the public cannot evaluate whether these decisions serve the broader public interest or primarily institutional interests.
2
Trust as currency in AI policy debates Both OpenAI and government institutions are asking for institutional trust precisely when trust in these institutions is eroding. Without corresponding transparency, oversight, or democratic input, requests for trust appear as power plays rather than good-faith appeals.
3
Military-AI partnerships reshaping tech landscape The deepening relationship between AI companies and defense establishments is reshaping how AI gets developed and deployed. This alignment means that national security imperatives—which often require secrecy—will increasingly drive AI policy away from public accountability mechanisms.