AI: Social network user does not control training data – 09/20/2024 – Tec

Social media companies collect, share and process vast amounts of information about their users, offering little transparency or control, including over how it is used by systems incorporating artificial intelligence, the US Federal Trade Commission (FTC) said in a report released on Thursday (19).

The report looked at how Meta, TikTok, Twitch and other social media companies manage user data, concluding that policies at many of these companies are “woefully inadequate.”

YouTube, X (formerly Twitter), Snap, Discord and Reddit were also included in the FTC report, although the findings in these cases were not released by the US agency.

Social media companies collect data through tracking technologies used in online advertising and by purchasing information from data brokers, among other means, the FTC said.

“While profitable for companies, these surveillance practices can jeopardize people’s privacy, threaten their freedoms, and expose them to a range of harms, from identity theft to stalking,” said FTC Chairwoman Lina Khan.

Big tech companies have been scrambling to acquire data sources to train their emerging artificial intelligence technologies. The data deals are often under-disclosed and often involve private content locked behind paywalls and login screens, with little or no notification to the users who posted it.

In addition to collecting data on how users engage with their services, most of the companies the FTC reviewed collected users’ ages and genders or guessed them based on other information. Some also collected information about users’ income, education and family status, the agency said.

The companies collected data on individuals who did not use their services, and some failed to identify all the ways in which data was collected and used, the FTC said.