Margrethe Vestager backs new AI Safety Institute for children after decade regulating Big Tech

A new independent AI Safety Institute for child protection will launch at the Danish Parliament, with Margrethe Vestager among the co-hosts.

Run by Common Sense Media, the institute will create crash-test-style evaluations for AI tools, offer open-source tests, and publish independent findings.

Funding comes from philanthropies and companies including Anthropic, the OpenAI Foundation, and Pinterest, while advisory-board conflicts will be tightly limited.

The launch highlights concerns that current EU and UK rules leave AI chatbots in a regulatory gray area, especially for youth mental health risks.