Microsoft CEO Satya Nadella has called for stronger safety measures in artificial intelligence systems, saying the industry needs to "step back and assess the trust architecture" of AI. In a post on social media, he outlined several safety ideas: separating AI models from the systems that control their work, keeping safety controls visible and external rather than hidden, and requiring that all significant actions be recorded in a way humans can read and verify.
Nadella emphasised that authorised people should always be able to pause or shut down an AI system while it is running, comparing this to an emergency brake. He stressed that developers should assume AI models might be compromised or behave unexpectedly and design systems to contain and control them from the beginning.
Why it matters
- Students and parents should understand that AI companies are actively discussing how to make systems safer and more controllable.
- Schools may need to consider AI safety principles when adopting AI tools for teaching and learning in classrooms.
- Hong Kong and Singapore families should monitor how local education authorities respond to global AI safety discussions and guidelines.
What to do
- Read your child's school's AI policy to understand what safeguards are in place for student data and AI use.
- Stay informed about AI developments by following announcements from major technology companies and regulators.
- Ask schools how they ensure AI tools used in classrooms have appropriate human oversight and safety controls.