Government & Enterprise AI News

All

1. OpenAI will honor a shared robots.txt signal for AI-training preferences, building on IETF working-group efforts

   Businesses with websites can state training preferences across vendors in one line, though whether opt-out truly protects publishers is disputed.

   @TheRealAdamG276 likes · 12 reposts · 31 replies · 85,475 viewsView post

   ExpandCollapseOpenAI will honor a shared robots.txt signal for AI-training preferences, building on IETF working-group efforts

   OpenAI's Adam G shared OpenAI's Oct 30 post: the company will honor a common robots.txt signal that lets websites tell multiple participating providers at once whether their content may be used for foundational generative-AI training. The post says it draws on the IETF AI Preferences working group, will be implemented in the coming months and won't replace existing crawler controls. The Verge (via Techmeme) reports Google, Microsoft and Anthropic will adopt the standard too. Critic Ed Newton-Rex calls such opt-out schemes “fake consent,” arguing training should require opt-in consent.

2. Axios via Techmeme: OpenAI plans a post-midterm policy and PR blitz, pushing mandatory “speed limits” for powerful AI and rebranding data centers as “AI compute campuses”Unconfirmed

   If data-center siting and “speed limit” rules take shape, they will affect compute supply, power prices and local investment policy.

   @Techmeme9 likes · 0 reposts · 3 replies · 3,307 viewsView post

   ExpandCollapseAxios via Techmeme: OpenAI plans a post-midterm policy and PR blitz, pushing mandatory “speed limits” for powerful AI and rebranding data centers as “AI compute campuses”

   Techmeme's summary of an Axios report says OpenAI plans a post-midterm policy and PR blitz, pushing mandatory “speed limits” for powerful AI and rebranding its data centers as “AI compute campuses.” Accounts such as kimmonismus joked that OpenAI is also rebranding AI as “superintelligence.” The underlying details — what the speed limits mean and any legislative path — could not be read here; this relies on the summary. It shows leading labs actively shaping the regulatory agenda and public perception of data centers.

3. The Information exclusive: OpenAI steps up efforts to block Chinese labs from extracting its models' reasoning as Astra becomes a major distillation targetUnconfirmed

   Anti-distillation measures may spill over onto legitimate bulk usage and data generation; teams relying on foreign APIs should watch term changes.

   @theinformation92 likes · 9 reposts · 13 replies · 12,515 viewsView post

   ExpandCollapseThe Information exclusive: OpenAI steps up efforts to block Chinese labs from extracting its models' reasoning as Astra becomes a major distillation target

   The Information reports exclusively that OpenAI says it is stepping up efforts to block Chinese labs from extracting its models' reasoning, even as Astra becomes a major distillation target. The full story is paywalled, so the specific techniques, labs named and evidence could not be verified here; this summary relies on the post's headline. Together with recent disputes over Anthropic flagging users for “training a competitor,” it suggests frontier labs are broadly tightening limits on how model outputs can be used.

4. AP: AI companies turn to religion to teach chatbots right from wrong; Anthropic says it has met 15+ religious and cross-cultural groups

   Where model values come from is now a public question; public-facing AI in service or health advice must be able to explain its ethical boundaries.

   @unusual_whales1,493 likes · 100 reposts · 247 replies · 233,662 viewsView post

   ExpandCollapseAP: AI companies turn to religion to teach chatbots right from wrong; Anthropic says it has met 15+ religious and cross-cultural groups

   unusual_whales relays an AP report that some AI companies are turning to religious traditions to shape chatbots' moral judgment. In its “Widening the conversation” post, Anthropic says it has met scholars, clergy, philosophers and ethicists from more than 15 religious and cross-cultural groups to inform Claude's constitution, trained values and evaluations, stressing it is not aligning with any single tradition. The post also describes an experiment giving Claude a mid-task tool that returns a reminder of its own ethical commitments, which Claude called right before consequential actions. AP also quotes experts warning that reducing diverse faiths to extractable rules can distort them.