The declaration to be unveiled as part of the Paris AI Action Summit is “not strong enough”, according to the artificial intelligence research group Ada Lovelace Institute.
The group warned the summit’s agenda lacked the “safety commitments of previous summits”, notably the AI Safety Summit held in the United Kingdom in 2023 that saw the signing of the Bletchley Declaration.
“To be the architects of a positive future for AI, governments need to build out the incentives required to make sure the technology works, is safe and is trustworthy – making sure the roof doesn’t fall in on us,” said the Institute.
“Governments also need to build and invest in alternatives that ensure the value and benefits of tech advances can be felt by everyone – and avoids paying extortionate rent to a few large companies for a generation.”
The research body noted that the Paris AI Action Summit’s declaration included no tools to ensure tech companies are “held accountable for harms” and warned of a “growing gap between public expectations of safety and government action to regulate”.
Announced during the initial AI Safety Summit, the Paris AI Action Summit is the third meeting of global political and technological leaders to attempt to reach international agreements on the future of artificial intelligence, the second being the AI Seoul Summit, ostensibly held in South Korea but which was largely attended remotely.
Politico reported on Monday that sources familiar with the matter claimed that the UK would not sign the diplomatic declaration that will be officially unveiled at the end of day one of the summit.