On Thursday, Meta said it is using advanced AI systems to enforce content policies. The company plans to rely less on third-party vendors. These systems will find and remove content related to terrorism, child exploitation, drugs, fraud, and scams.
Meta says it will roll out these AI systems across its apps once they consistently outperform existing methods, though no specific date has been given for the full rollout.
Meta explained in a blog post that people will still review content. However, the new systems will handle tasks better suited to technology. This includes repetitive reviews of graphic content and areas where bad actors often change their tactics. Examples are illicit drug sales or scams.
Meta believes these AI systems can more accurately identify violations, help prevent scams, respond faster to real-world events, and avoid over-enforcement.
For example, according to the company, early tests show the AI systems can detect twice as much adult sexual solicitation content as review teams. They cut the error rate by over sixty percent. The systems can spot and prevent more impersonation accounts involving celebrities and other high-profile people. They helped stop account takeovers by detecting activities such as logins from new locations, password changes, or profile edits.
Meta says its systems can stop about 5,000 scam attempts each day, in which scammers try to obtain users’ login information.
Meta wrote in the blog post that experts will design, train, oversee, and evaluate our AI systems, measuring performance and making the most complex. High-impact decisions. The company added that people will still play a key role in the highest risk and most critical decisions, such as Appeals of account disablement or reports to law enforcement.
Over the past year, as President Donald Trump began his second term, Meta has relaxed its content moderation rules. Last year, the company replaced its third-party fact-checking team with a Community Notes model similar to X. More recently, Meta removed restrictions on topics considered part of mainstream discussion and announced that users will be encouraged to take a more customized approach to political content.
At the same time, Meta and other major tech companies are facing lawsuits that aim to hold them responsible for harming children and young users.
Also on Thursday, Meta announced it is launching a Meta AI support assistant. This assistant, available beginning immediately, provides users with 24/7 help. It will be available worldwide on the Facebook and Instagram apps for iOS and Android, as well as the help center on Facebook and Instagram for desktop users.
Source: Meta rolls out new AI content enforcement systems while reducing reliance on third-party vendors
