Stenobird

Episode

At the Pentagon, OpenAI is In and Anthropic Is Out

Podcast
Hard Fork
Published
Mar 1, 2026
Duration seconds
1988
Processing state
processed
Canonical source
https://www.nytimes.com/column/hard-fork
Audio
https://dts.podtrac.com/redirect.mp3/pdst.fm/e/pfx.vpixl.com/6qj4J/pscrb.fm/rss/p/nyt.simplecastaudio.com/3e43d072-f8a5-430f-bc8e-4c70aafdf3c7/episodes/9d2e9763-11ae-4556-b30b-bbc50b5323d2/audio/128/default.mp3?aid=rss_feed&awCollectionId=3e43d072-f8a5-430f-bc8e-4c70aafdf3c7&awEpisodeId=9d2e9763-11ae-4556-b30b-bbc50b5323d2&feed=l2i9YnTd
JSON
/v1/public/podcasts/hard-fork/episodes/at-the-pentagon-openai-is-in-and-anthropic-is-out
Markdown
/podcast/hard-fork/at-the-pentagon-openai-is-in-and-anthropic-is-out.md

Actions

  • POST https://stenobird.com/v1/public/podcasts/hard-fork/episodes/at-the-pentagon-openai-is-in-and-anthropic-is-out/transcription-requests
    Idempotently request low-priority transcript generation for this episode.
  • GET https://stenobird.com/podcast/hard-fork/at-the-pentagon-openai-is-in-and-anthropic-is-out.md
    Read the agent-friendly Markdown representation of this episode resource.

Summary

The Pentagon's sudden pivot from rejecting Anthropic to partnering with OpenAI signals a high-stakes shift in government AI procurement. This episode explores whether the move is driven by substantive safety differences or unprecedented political pressure on tech firms.

Topics

  • OpenAI
  • Anthropic
  • Pentagon
  • Artificial Intelligence
  • Government Contracting
  • Regulatory Capture
  • National Security
  • Tech Policy

Highlights

  • Main idea: The US government's rejection of Anthropic's terms while accepting OpenAI's suggests a potential move toward regulatory capture
  • Failure mode: Using government contracts as a tool to punish companies based on ideological differences could mirror authoritarian tech regulation
  • Practical takeaway: The lack of clear legal frameworks for AI usage in military applications creates significant risks for domestic surveillance
  • Main idea: The suddenness of OpenAI's agreement suggests a strategic maneuver to align with federal interests during a period of high volatility
  • Failure mode: Relying on 'legalese' and vague promises of safety rather than enforceable laws may fail to prevent misuse of powerful AI systems

Chapters

  1. 1:00 The 48-Hour AI Chaos: An introduction to the rapid and confusing developments involving the Pentagon, Anthropic, and OpenAI.
  2. 3:50 The Anthropic-Pentagon Breakdown: Details on the breakdown of negotiations between the Pentagon and Anthropic's leadership.
  3. 6:55 The Defense Production Act Question: An analysis of why the failure to invoke the Defense Production Act was a significant signal for Anthropic.
  4. 9:25 OpenAI's Sudden Agreement: Examining Sam Altman's announcement regarding OpenAI's new deployment agreement with the Pentagon.
  5. 11:55 Risks of Military AI: The dangers of deploying powerful AI systems in military contexts without established legal safeguards.
  6. 17:20 Ideological Regulation: A discussion on the chilling effect of the government using contracts to enforce ideological alignment.
  7. 27:20 The Reality of Regulatory Capture: Evaluating if OpenAI is successfully leveraging political influence to navigate government disputes.