Newsletters
Microsoft has introduced mandatory sign-in requirements for its Copilot AI assistant and restricted access for children under 13 as part of a new online safety initiative unveiled Thursday.
The company's Safe Participation Framework rests on three principles: safety by design, age-differentiated experiences, and education and empowerment, Microsoft said in a statement.
Under the age-differentiated pillar, all users must now sign in to access Copilot, while children under 13, or a higher age threshold where required by local law, face restricted access to the assistant. Microsoft said the requirement is intended to tailor experiences to a user's age rather than apply a single set of rules to all users.
On safety by design, Microsoft said the approach identifies foreseeable risks, evaluates the impact of product features, and applies safeguards proportional to those risks before releasing products to the public.
As part of the education and empowerment pillar, the company said it has added measures to reduce user exposure to harmful content, direct people to crisis-support resources when it detects signs of self-harm risk, and encourage healthy usage patterns through break reminders and reporting tools. It also introduced safeguards intended to limit delusional or misleading AI-generated outputs.
Microsoft's Bing search engine has strengthened protections against child sexual abuse material within its generative AI services and expanded its SafeSearch feature globally to cover additional categories of harmful content, the company said.
Separately, Microsoft introduced the Windows Age application programming interface (API), which gives developers access to privacy-preserving age signals so they can build age-appropriate experiences without collecting or sharing unnecessary personal data.
In its statement, Microsoft said, "Children and families should not have to choose between the protections they need and the opportunities technology can provide."
The company said technology alone cannot guarantee safe digital participation and that it plans to deepen cooperation with researchers, AI safety institutes, governments and civil society organizations over the coming year.
Set as preferred source
Set as preferred source
2 min read
