Today's Headlines

  • OpenAI, Google and Microsoft will honor a shared signal for opting out of AI training — one robots.txt entry instead of one per company
  • Google's prototype satellite returns the first AI response from a TPU in orbit — twin linked satellites are next, in 2027
  • OpenAI tightens weapons information for teens — and will publish a community safety report every six months

Today's theme is signals. Websites that want to keep their pages out of AI training have had to write separate rules for each AI company's crawler. The largest AI companies are now lining up behind a single signal.

Google heard back from an AI chip it placed in orbit, and OpenAI is giving ChatGPT stronger responses to warning signs of violence in conversations.

Today's Top Three

OpenAI, Google and Microsoft will honor a shared signal for opting out of AI training

OpenAI, Google and Microsoft each said on October 30 that they will honor a shared signal websites can use to keep their content out of AI training.

The signal is a common "policy token" placed in a site's robots.txt file. It draws on the work of the IETF's AI Preferences working group, and Kevin Kelley of Anthropic submitted an Internet-Draft describing it the same day. Until now, site operators have had to name each AI company's crawler and write a separate rule for it.

Timing differs by company. OpenAI says it will implement the signal in the coming months, and Google aims to support it alongside other providers in 2027. Microsoft will honor the new preferences as it crawls starting February 1, 2027. According to The Verge, Anthropic has already adopted the signal.

Company-specific controls stay in place. Google will read an opt-out through the shared token as a disallow of Google-Extended when a site has not set Google-Extended explicitly. Microsoft is adding a separate page-level preference for grounding, the use of web content to inform generated answers, and says sites remain eligible for Bing search results either way.

Google says the approach is consistent with the EU AI Act's Code of Practice for general-purpose AI model providers, which treats robots.txt as a primary way for publishers to express machine-readable preferences.

Google's prototype satellite returns the first AI response from a TPU in orbit

Google said on October 30 that it has received the first AI response from the prototype satellite for Project Suncatcher, its research effort to run AI compute in space.

The satellite launched on October 1, and Google spent the following weeks powering up the host and TPU trays and running early machine learning tests. During a ground-station pass, the team sent the onboard Gemma 4 open model a prompt asking for a haiku about being in space, and the reply came down as the satellite passed over Svalbard, Norway. Google calls it the first AI response from a data-center-class TPU in orbit.

Next come tests of how the TPUs perform under thermal stress and radiation. Once the chips prove they can run workloads reliably in orbit, Google plans to move to twin interconnected satellites in 2027.

OpenAI tightens weapons information for teens and commits to a twice-yearly safety report

OpenAI said on October 30 that it is strengthening violence prevention in ChatGPT and further limiting the weapons information users under 18 can receive.

In the teen experience, ChatGPT will restrict practical information about weapons and security systems that could be misused, even when a question shows no sign of violent intent. OpenAI has also strengthened age prediction so that it keeps assessing new signals and moves likely teens into the under-18 experience. Adults placed there by mistake can confirm their age through Persona, OpenAI's age verification partner.

Policies for all users are tighter as well. ChatGPT will restrict reconstructions of past mass attacks, including the weapons and equipment used, how they were obtained, and detailed comparisons of the scale of harm, regardless of a user's age or reason for asking. When a conversation shows signs that someone is fixating on violence, ChatGPT will be able to end the chat, and accounts with repeated endings may be paused temporarily.

OpenAI says it made more than 200 referrals to law enforcement in the third quarter of 2026. It will publish its first Community Safety Report in early 2027, covering the second half of 2026, and update it every six months after that. The new safeguards will roll out over the next few weeks.

Other Developments

Models & APIs

Research & Safety

Policy & Law

Business

Source: selected by the editors from the AI news inbox (collected October 31, 2026: 32 items, 5 primary and 27 secondary).