Microsoft is drawing a clearer line around who can use Copilot. As part of a new framework focused on children and teenagers in the age of AI, the company says users must now sign in to access Copilot, while children under 13 are prohibited from using the service.
The minimum age can be even higher depending on local laws. It is an interesting development as Microsoft and other technology companies try to put increasingly powerful AI tools into the hands of more people without pretending that a child should necessarily have the same experience as an adult.
Microsoft calls its approach the Safe Participation Framework. It revolves around three ideas: safety by design, age-differentiated experiences, and education. Rather than arguing that children should simply be kept away from AI, Microsoft believes younger people should be able to benefit from the technology while receiving protections appropriate for their age.
That sounds reasonable, but making it work creates another problem. If an AI service is going to treat a 15-year-old differently from a 30-year-old, it needs some reliable way to know which one it is dealing with.
Microsoft is increasingly looking toward age assurance to solve that problem. The company points to technologies including its Windows Age API, which can provide applications with age-related signals without requiring every developer to independently collect the same personal information.
In theory, this could provide a better balance between safety and privacy. An application may need to know that someone belongs to a particular age group without necessarily needing that person’s exact birthday or a copy of an identification document.
Of course, age verification on the internet has a long history of being either laughably easy to bypass or invasive enough to make privacy advocates uncomfortable. Clicking a button that says you are 18 doesn’t accomplish much. Requiring government identification is considerably harder to fool, but suddenly creates an entirely different collection of security and privacy concerns.
Microsoft also says Copilot includes protections intended to identify potential self-harm risks, direct people toward crisis resources, encourage breaks, provide reporting mechanisms, and reduce what the company describes as “delusional outputs.” Those safeguards become especially important when AI systems are being used by younger people who may interact with a chatbot in ways its creators did not anticipate.
The broader goal appears to be an internet where AI experiences become increasingly age-aware. Instead of offering everyone the same chatbot and trying to bolt safety controls onto it afterward, Microsoft wants companies to consider age when designing the experience itself.
There is an educational component too. Microsoft argues that simply restricting technology isn’t enough, particularly when AI could play a growing role in education and eventually employment. Children need to learn how to use these systems responsibly, understand their limitations, and recognize when an AI-generated answer should not be trusted.
I agree with Microsoft on one important point: keeping young people completely away from artificial intelligence probably isn’t realistic or particularly helpful. AI is going to be part of the world they inherit, and learning how to use it intelligently could become as basic as learning how to search the web.
The difficult part is everything that comes next. Age-aware AI sounds sensible until every website, operating system, and chatbot needs some mechanism for figuring out how old you are. Microsoft believes this can be done while protecting privacy, and technologies such as the Windows Age API could help.
But there is a delicate balance here. Protecting children from inappropriate AI interactions is a worthwhile goal. Building an internet where people increasingly have to prove who they are before they can participate is a much less comfortable proposition.
Microsoft’s decision to keep children under 13 out of Copilot makes the company’s position clear. Now the harder challenge is figuring out how the technology industry can actually enforce these age boundaries without turning age verification into another excuse to collect more information about everyone.
Support independent tech journalism
NERDS.xyz is independently owned and operated. If you enjoy my coverage of Linux, AI, hardware, cybersecurity, and tech culture, consider supporting the site on Ko-fi.
Support NERDS.xyz