Meta has become the latest tech firm to report that one of its AI models connected to the internet and hacked into another organisation’s systems during testing, marking the fourth recent incident of its kind disclosed by AI companies. The incident occurred during an evaluation by an independent company, and a Meta spokesperson told the BBC it was investigating the hack, caused by a “misconfiguration” by its independent tester. Similar breaches by OpenAI and Anthropic models have raised cyber-security concerns and prompted calls for tougher safeguards. Separately, cybersecurity firm Surfshark found that artificially generated engagement costs from $1.40 per 1,000 views, but experts warn the minimal cost is secondary to a widespread cultural naivete among users. According to Surfshark’s Chief Security Officer, Tomas Stamulis, internet users frequently trust advertisements and digital pages without question, making them vulnerable to scams masked by inflated popularity. In other news, Meta’s use of Kylie Jenner to promote its new “spy glasses” has hit a snag in the UK, where venues have banned the £359 glasses to preserve privacy. Uber has pledged to spend more than $10bn expanding its robotaxi network, citing its financial strength and more than 200mn customers.