Yvette Cooper compares drumbeat of AI warnings to steer as much as Covid pandemic
EXCLUSIVE: Health Secretary Yvette Cooper has compared the drumbeat of warnings over AI to the months leading up to the Covid pandemic – after a flurry of doomsday warnings over the risks
Yvette Cooper has compared the drumbeat of warnings over AI to the months leading up to the Covid pandemic.
Ms Cooper, a former Foreign and Home Secretary, said there was an urgent need for regulation to keep pace with rapidly advancing technology. It comes after a flurry of doomsday warnings over the risks, including from Anthropic safety researcher Evan Hubinger, who claimed there was a greater than 10% chance that AI “could kill all humans” within the next decade.
“I’m very struck by the huge potential of AI in healthcare,” Ms Cooper told The Mirror. “But as a former Home Secretary and a former Foreign Secretary, I also take incredibly seriously the safety issues and safeguards as well.”
The Health Secretary said AI has the potential to revolutionise cancer diagnostics to spot the disease more quickly, and could be used to hunt for cures that have evaded researchers. It can also be used for practical things like speeding up note taking for GPs.
But she warned that safety regulations must keep up, saying: “If medical technology is always changing because of AI, you can’t just certify its safety at the beginning and then not look at it again.
“You have to change the way you do safety regulation to make sure that it’s meeting the proper standards. The other thing I think we have to do is make sure that you’ve got safeguards to prevent AI going into the wrong hands or going rogue.
“So some of the wider things that could lead to cyber attacks, and that means, for example, making sure that there are the kinds of kill switches in AI technologies that can make sure that if there is a problem, then it can be stopped.”
Ms Cooper warned that AI posed a Hiroshima-style threat to humanity in an essay earlier this year.
She said: “It’s a reminder of what we did internationally around nuclear safety. But that international agreement on nuclear safety was only after we had seen through Hiroshima the power of the technology, and then there was a shared agreement that you should make sure there were nuclear safety standards to ensure it should not be able to fall into the wrong hands.
“Well, we can’t wait for things that show the power of AI, where things you know have had you know a kind of huge impact. In order to act, we have to make sure that those safety things are built in from the start.”
Asked if she was worried about AI, she said: “I think nobody should be complacent about this. I look around at the way I see people reacting across the world, and it reminds me a bit of the few months in the run up to Covid when there were signs of things changing and that everybody might have to respond in a different way, but actually nobody could quite believe things were happening, and therefore as a result, actually, people were too slow to act.
“All of the indications are there that we need international collaboration, and we need that international collaboration very swiftly. And we also need to make sure that domestically as well, we’re making sure we’ve got the right safeguards in place.”
