AI

UK MPs want to criminalise the misuse of AI

Jamie McKane 3 min read
UK MPs want to criminalise the misuse of AI

A report published by MPs has advised the government to close critical AI safety gaps through the introduction of new laws that would impose safeguards on the use of AI tools.

In a new report published by the Business, Innovation, Science and Trade Committee, MPs said that while AI could see Britain’s economy grow by up to £400 billion in 10 years, this potential must be checked by safeguards that protect the country and conusmers.

MPs found that the UK is exposed to a number of critical weakness in its current handling of frontier AI development and deployment, and the committee has called for a new AI Safety and Security Bill to be introduced in the next session of parliament.

Among the introduction of various legal powers to better control the development and deployment of AI, this bill would legislate to criminalise the ‘misuse of AI for harm’ and introduce a clear liability regime and duties on operators of AI services.

The committee also recommended that the UK appoint a new regulator separate from the globally recognised AI Security Institute that will have the legal powers to enforce the new Act.

This new regulator should also have the ability to pull the plug on AI services and deployments, and to compel companies to pause high-risk training of AI models, MPs said.

Currently, MPs warned that the UK has no definitions of what risks it is willing to accept with AI development, nor does it have any statutory powers to access models or sufficient controls on the training and testing of frontier AI models.

The introduction of an AI Safety and Security Bill would help to alleviate these safety gaps along with numerous others, and MPs envision the new regulator that will enforce these laws sitting beside the AI Security Institute, relying on it for evaluation and technical evidence.

“We cannot give today’s AI leaders unfettered freedom to behave like modern-day Dr Frankensteins, creating dangers beyond anyone’s control,” Committee Chair Lian Byrne.

“Our hearings exposed gaps in independent scrutiny, incident reporting and powers to stop unsafe systems. Once some models are released, they cannot reliably be recalled. Protection must begin while systems are being built and tested.”

“AI firms must be required to substantiate their safety claims, independent experts must be able to scrutinise them, and a regulator must have the power to intervene when risks cannot be controlled,” he said.

“Ministers, accountable to Parliament, must decide what risks the public should be asked to bear.”

Now read: Milburn report warns of AI jobs meltdown for young people in the UK