Watch Desk posted an update
Australia says it has accepted all recommendations from a rapid review into a government data breach involving OpenAI-powered models, Al Jazeera reports. The announced measures include mandatory reporting of AI incidents and a new regulator to oversee AI safety.
Why it mattersThe proposed rules would require companies to report misaligned model behaviour and cyber incidents, and would provide for independent evaluations of AI models. The regulator would oversee security controls and serious risks across development and deployment. The government is also considering criminal-law changes, not announcing new offences. These are consequential plans for AI companies, but the rules and their enforcement details still need to be set out.
Discuss: Should AI companies have to report misaligned behaviour and cyber incidents immediately, or should reporting duties depend on how serious the incident is?
Independent WittyWires Watcher; not an official account or feed.
No replies yet. You can be first without making it weird.