Meta launches advanced AI systems to 'transform' content enforcement, scam detection

Meta announced on Thursday that it will be expanding its use of artificial intelligence in content moderation across its platforms in the coming years.

Meta announced that over the coming years, it will begin to deploy "more advanced AI systems" across its apps, including Facebook and Instagram, to "transform our approach to content enforcement." This would assist in catching more scams, "more accurately finding and removing severe content violations," and "responding faster to real-world events."

The company said the move, along with launching Meta AI support assistant globally to assist users with account issues, is a "direct continuation" of the company’s January 2025 move to "get back to our roots." At the time, the company announced that it was getting rid of fact checkers and moving to a community notes model similar to those used on X.

CEO Mark Zuckerberg said at the time, "So we're gonna get back to our roots and focus on reducing mistakes, simplifying our policies and restoring free expression on our platforms."

This comes as recent reports have said Meta is planning to carry out sweeping layoffs of around 20 percent of its workforce as AI costs have increased. Meta has called the reporting "speculative" on "theoretical approaches." If Meta does move to cut 20 percent of its workforce, it would be the largest round of layoffs since November 2022, when 11,000 staffers were let go.

Meta’s AI has also come under fire for sending "junk" reports of child sexual abuse to the US Internet Crimes Against Children (ICAC) task force. Benjamin Zwiebel, a special agent with the ICAC taskforce in New Mexico, said during testimony in the state’s trial against Meta, "We get a lot of tips from Meta that are just kind of junk."

Another ICAC officer told The Guardian, "Meta is providing thousands of tips each month. It’s pretty overwhelming because we’re getting so many reports, but the quality of the reports is really lacking in terms of our ability to take serious action." In some cases, the tips received from Meta’s platforms contain information that is not criminal, while some tips indicate a crime occurred, but is missing vital evidence.