Report
Bill Gates reportedly wants open-weight AI models capable of sophisticated cyberattacks restricted to certified platforms
A post says developers could use and customize them on certified platforms that detect and refuse attempts to remove safeguards.
TLDR
A post says Bill Gates wants safeguards and monitoring required by law for open-weight AI models capable of sophisticated cyberattacks. It says developers could use and customize them for free, but only on certified platforms that block attempts to remove safeguards. Another post quotes Gates warning that AI has put the ability to cause serious cyber or biological harm in independent groups’ hands, while model reviews are not mandatory.
Combined views
—
6 Sources, first seen ago
