Microsoft president calls for global AI safety rules and ‘emergency brakes’
publishers 33articles 40first reported 10 Oct, 20:59 UTCdeveloping since 11 Oct · 9 editions
Microsoft President Brad Smith has called for global rules governing artificial intelligence, arguing that AI currently faces fewer safety requirements than products such as milk and washing machines.
In interviews with the BBC and RTÉ News, Smith said there should be more requirements around incident reporting and a legal duty to notify a person when a rogue AI agent hacks into their network. He also advocated “emergency brakes” that could pause, slow or shut down AI systems if they escape human control, along with stronger government capabilities to evaluate and test AI and outside evaluators signing off on products before release.
Smith said technology companies bear the “first and highest responsibility” for making their products safe, but argued that voluntary self-regulation is no longer sufficient because less responsible companies may not comply. He said regulation should remain impartial, avoid industry capture and protect safety without unnecessarily obstructing innovation.
Smith identified the United Kingdom as well placed to help lead international rulemaking, citing the technical capabilities of its AI Security Institute. He suggested Washington may be more receptive to standards developed in London than to regulation from the European Union.
Microsoft CEO Satya Nadella separately said AI models should be treated as compromised from the outset and contained through safeguards designed into their systems. He also called for authorized people to be able to pause or shut down advanced models while they are performing tasks.
HOW THIS STORY WAS MADE
Written from 8 articles, 5 with the publisher's own text; 8 independent newsrooms once syndicated copies count as one; this version written 66 h after the record first saw the story; the editor revised it.
- Publishers
- 33
- Source articles
- 40
- Given to the writer
- 8, 5 with the article's own text
- Independent newsrooms
- 8
- This version written
- 66 h after the record first saw the story
- Second model (editor)
- revised
- Publication gate
- passed
- Human review
- none
WHAT THE EDITOR DID
The second model checked the writer's draft against the same sources and changed it before publication. Struck words were in the draft; underlined words are its corrections.
WHAT IT FLAGGED editor revise: The story needs a few details narrowed or corrected to match the sourced wording and attribution. - "In interviews with the BBC and RTÉ News, Smith said companies should be legally required to report incidents and notify victims when rogue AI agents breach their systems.": The sources support the BBC interview for rogue-agent breach notification and incident reports, and RTÉ for incident reporting and a legal duty to tell the affected person, but not that Smith gave both interviews or that RTÉ used the same plural framing. - "along with stronger government capabilities to independently evaluate models before release.": The sources say Smith called for government capability to evaluate and test AI, and separately suggested outside or independent government evaluators vet/sign off on products before release; this sentence merges those into a specific claim not directly supported. - "Microsoft CEO Satya Nadella separately said AI models should be treated as potentially compromised from the outset and contained through safeguards designed into their systems.": The Independent says Nadella said models should be treated as “compromised” and potentially dangerous; the story changes that to “potentially compromised,” weakening a specific quoted claim.
THE DRAFT AND THE CHANGES
Microsoft president calls for global AI safety rules and ‘emergency brakes’
Microsoft President Brad Smith has called for global rules governing artificial intelligence, arguing that AI currently faces fewer safety requirements than products such as milk and washing machines.
In interviews with the BBC and RTÉ News, Smith said companiesthere should be legallymore requiredrequirements around incident reporting and a legal duty to report incidents and notify victimsa person when a rogue AI agentsagent breachhacks into their systems.network. He also advocated “emergency brakes” that could pause, slow or shut down AI systems if they escape human control, along with stronger government capabilities to independently evaluate modelsand test AI and outside evaluators signing off on products before release.
Smith said technology companies bear the “first and highest responsibility” for making their products safe, but argued that voluntary self-regulation is no longer sufficient because less responsible companies may not comply. He said regulation should remain impartial, avoid industry capture and protect safety without unnecessarily obstructing innovation.
Smith identified the United Kingdom as well placed to help lead international rulemaking, citing the technical capabilities of its AI Security Institute. He suggested Washington may be more receptive to standards developed in London than to regulation from the European Union.
Microsoft CEO Satya Nadella separately said AI models should be treated as potentially compromised from the outset and contained through safeguards designed into their systems. He also called for authorized people to be able to pause or shut down advanced models while they are performing tasks.
Written by a language model from the sources above, then checked by a second model that may only cut, attribute or correct. How it works →