In brief
- OpenAI said it will follow a shared signal that lets websites communicate their AI training preferences to multiple providers.
- The signal will be included in robots.txt, the file websites use to specify crawler preferences.
- OpenAI aims to implement the shared signal in the coming months but has not given a firm date.
- The company says existing robots.txt crawler controls will remain in effect alongside the new signal.
In an announcement on October 30, 2026, OpenAI said it would follow a shared signal allowing website operators to communicate their preferences about the use of their content for generative AI training to multiple providers. The company aims to implement this approach through robots.txt in the coming months. OpenAI says it will share details of its implementation later.
What does the shared signal offer site administrators?
Website administrators may have to communicate their AI training preferences separately to different providers. OpenAI's announcement says a shared signal would make it easier to communicate the same preference to multiple participating providers and reduce the burden of maintaining separate rules for each crawler.
The signal will be placed in robots.txt, where preferences related to crawlers can be specified. OpenAI describes the effort as a voluntary, collaborative initiative and encourages other AI providers to follow the signal. The announcement is connected to work underway in the Internet Engineering Task Force (IETF) AI Preferences working group.
OpenAI's announcement does not provide a technical example of how to write the signal or a firm implementation date. The company says it will share implementation details later, so it cannot be concluded that the shared approach currently works the same way across all providers.
Providers are working to different timelines
In its October 30 report, The Verge describes the shared method as a “policy token” signal that different providers can recognize. According to the report, Anthropic has started using the signal, while Google and Microsoft plan to support it next year. This timeline information comes from The Verge; OpenAI's own stated target for implementation is in the coming months.
The Verge also reported that OpenAI, Google, Microsoft, and Anthropic would restrict access to the relevant sites by AI crawlers that detect the signal. It also reported that Meta had expressed support for the shared signal, but noted that the company had not made a clear commitment about whether it would follow it. So an expression of support is not the same as actually implementing the signal.
Existing robots.txt controls will remain in place
OpenAI specifically says the shared signal will not replace existing robots.txt controls. The company's current crawler rules will continue to apply. This means the announcement does not indicate that previous controls have been removed or that website owners need to configure anything new right away.
When implementation details are published, the signal's format and which providers recognize it will become clear. For now, OpenAI's announcement is limited to its decision to support the shared approach and its target of implementing it in the coming months. The announcement also does not comprehensively explain how it affects uses of site content beyond AI training.
What it means for architecture and visualization publishers
Architecture firms, interior designers, visualization studios, and those producing educational content may want to manage their preferences about using online project articles, images, and portfolio content for AI training. To the extent that supporting providers recognize it, the shared signal could reduce the need to set separate rules for each provider.
However, the approach's impact depends on which companies implement the signal and how they interpret it. Teams running project archives or online portfolios would be wise to keep their existing robots.txt controls in place until new technical details are announced, then check the signal's scope and provider support. The announcement does not introduce a new rule that changes licensing terms or content ownership.
Sources
1 sourceSource texts are not republished; short quotes are marked, everything else is our own summary and commentary.
For teams in Turkey publishing project portfolios, educational content, or visualization work, the shared signal could make it easier to manage preferences for AI crawlers. But its practical value depends on how many providers actually recognize the signal and how they implement it; the announcement does not guarantee control over all uses of content.
Implementation details have not yet been announced, so there is no need for firms to make a new technical change right away. Keeping existing robots.txt rules in place and monitoring providers' timelines is a more cautious approach. It is also important to assess licensing and content policies separately.
Frequently asked questions
When will OpenAI implement the shared robots.txt signal?
OpenAI said it aims to implement the signal in the coming months. It has not given a firm date and said it will share implementation details later.
Will the shared signal replace existing robots.txt rules?
No. OpenAI says the new signal will not replace existing robots.txt crawler controls, which will remain in effect.
What will the shared robots.txt signal do?
The goal is to let website administrators communicate their AI training preferences to multiple participating providers through a shared signal, reducing the burden of managing separate rules.
