AI will soon be powerful enough to "kill many humans", adviser warns

The Prime Minister’s adviser on Artificial Intelligence (AI) has warned that AI systems might be powerful enough to “kill many humans” within just two years, and has suggested countries need to work together to create regulations for its development.

His warning comes amid many headlines in recent weeks that AI is progressing too quickly and calls for a 6-month moratorium on progress to give time for research. Just last week 350 AI experts, including the CEO of OpenAI, which developed ChatGPT, signed a letter warning that it could lead to human extinction.

Matt Clifford, speaking on TalkTV said AI led to increased threats from cyberattacks and the creation of bioweapons: “You can have really very dangerous threats to humans that could kill many humans, not all humans, simply from where we’d expect models to be in two years time.”

“If we try and create artificial intelligence that is more intelligent than humans and we don’t know how to control it, then that’s going to create a potential for all sorts of risks now and in the future…it’s right that it should be very high on the policymakers’ agendas.”

The UK Government recently met with AI leaders to learn about the dangers of new technology; in recent years, leaders in the AI industry have become less optimistic about whether it is likely to benefit humanity.

Clifford warned that “the people who are building the most capable systems freely admit that they don’t understand exactly how they exhibit the behaviours that they do”.

The UK government has been keen to encourage innovation and Rishi Sunak has been quite public about his interest in using new technologies to drive forward the economy, but reports in the Times also suggest that he might also be pushing the UK forward as a centre for global regulation for AI.