A collective recourse against LinkedIn, a subsidiary of Microsoft, is currently making a sensation in the United States. Premium users (i.e. paid users with subscription) accuse the platform of transmitting private messages to third-party companies without explicit authorization in order to train AI models.
As the BBC reportsLinkedin would have “discreetly” introduced a new confidentiality parameter in August 2024 which automatically included user accounts for Data sharing for AI training (“disengage”). A month later, According to Reuters, A discreet adjustment was made to the Privacy Policy, which said for the first time the use of data for AI training purposes.
What is particularly controversial is that the FAQ adds later that a subsequent unsubscription has no influence on the use of data already carried out. The complainants see it as an attempt to “blur the traces”, as the court documents show.
Linkedin meters: “False -based statements”
The trial was filed in California, On behalf of LinkedIn Premium users. The company claims $ 1,000 of damages per user concerned for alleged violations of the US law on stored communications, as well as other amounts not specified for breach of contract.
Linkedin, which has more than a billion users worldwide-including almost a quarter in the United States-firmly denies these allegations. A company spokesperson told the BBC: “These are false baseless affirmations. “
The geographic component is interesting: as shown in an email from LinkedIn to its users, the sharing of data for AI purposes has not been activated in Britain, in the European Economic Area and in Switzerland. Users of these countries are therefore probably not affected by this case.
The stake is important for LinkedIn: the company has generated turnover of $ 1.7 billion from premium subscriptions Alone in 2024. The platform claims to know a rapid growth in the number of premium subscribers, which is also linked to the growing integration of AI’s features.