Surveillance Concerns Emerge as Critics Question OpenAI Deal

[johnblockbuster]

Despite OpenAI’s assurances, critics argue the Pentagon contract could still open the door to surveillance risks. Techdirt’s Mike Masnick claimed the agreement permits data collection compliant with Executive Order 12333 — a directive he described as enabling intelligence agencies to gather communications by tapping infrastructure outside the U.S., even when involving Americans.

Meanwhile, OpenAI’s head of national security partnerships, Katrina Mulligan, pushed back on the criticism, arguing that deployment architecture — not just contract language — prevents misuse. Altman acknowledged the backlash has been significant, with Anthropic’s Claude briefly overtaking ChatGPT in the Apple App Store rankings. Still, he defended the decision as a strategic move aimed at de-escalating tensions between government and the AI industry — a gamble that could either redefine OpenAI’s role in national security or deepen scrutiny over its priorities.

[madmax]

nothing calms surveillance fears like citing a cold war era directive