Will Meta be Allowed to Train its AI on User Data?; UK Retailers Launch Biggest Ever Collective Action; California Protests Proposed AI Bill

by Aimee Newell Tarín on 10th Jun 2024 in News

On today’s news digest: Will Meta be Allowed to Train its AI Models on User Data?; UK Retailers Launch Biggest Ever Collective Action; California Protests Proposed AI Bill

Meta has been hit with 11 EU complaints over proposed changes which would see it using users' data from its platforms to train its AI models without requesting consent. The recent changes in Meta’s privacy policy – which are due to come into effect on 26th June – would allow it to train models on users’ posts and online tracking data. This would likely breach privacy rules in the European Union. An advocacy group, NOYB, has urged privacy watchdogs within the region to halt the plans.

Amazon is also under fire for alleged data misuse. UK retailers have fined a £1.1bn damages claim against the e-commerce giant, alleging that it illegally misused their data for competitive purposes and to benefit its own commercial operation. This marks the biggest ever collective action undertaken by UK retailers. Allegedly, the data being used by Amazon belonged exclusively to the retailers on its marketplace. The claim will be filed on behalf of the retailers by the British Independent Retailers Association in London’s Competition Appeal Tribunal.

Things are also heating up in California, where AI industry heavyweights are protesting against a proposed state bill which would force tech companies to adhere to a strict safety framework regulating the technology. Currently OpenAI, Anthropic, and Cohere – some of the biggest companies leading innovation in the industry – operate within the state. Already passed by the California Senate last month, the Safe and Secure Innovation for Frontier Artificial Intelligence Systems Act requires AI groups in the state to give their guarantee to a new state body that they will not develop models with “hazardous capability”. Under the proposed regulation, developers would have to report on their safety testing and introduce a “kill switch” to shut down their models. Critics have accused the bill of being overly restrictive, and say it will drive AI leaders to relocate. The bill is now set for a vote from its general assembly in August.