Meta is launching its AI-powered support assistant globally across Facebook and Instagram, offering users round-the-clock help with account problems and marking a significant expansion of the company's use of artificial intelligence to police its platforms.
The assistant, previewed in December, is available through the iOS and Android apps for both platforms and through the Help Center on desktop, and can handle a growing range of requests directly within Facebook, with Instagram support to follow.
Tasks within scope include reporting scams and impersonation, adjusting privacy settings, resetting passwords and updating profile information, with responses typically delivered in under five seconds.
The rollout covers all languages supported by Facebook and Instagram for help topics, with limited login-assistance trials already underway in the United States and Canada.
The announcement was accompanied by enforcement figures that illustrate both the scale of Meta's content moderation challenge and the degree to which AI systems are beginning to outperform the human review processes they are gradually replacing.
More advanced AI enforcement systems currently being tested are identifying and acting on 5,000 scam attempts per day that previous review teams failed to catch, the company said.
Reports of the most impersonated celebrities have fallen by more than 80%, while the systems are detecting twice as much violating adult sexual solicitation content as before, with error rates falling by more than 60% at the same time.
Views of advertisements containing scams or other serious violations have dropped 7%, and the language coverage of the enforcement systems has expanded to cover languages spoken by 98% of people online, up from around 80% previously.
Related reading
- From Vera Rubins, Blackwell, Nemotron, NemoClaw, to Isaac and Cosmos - here's a whistlestop guide to NVIDIA's…
- Samsung and AMD sign memory partnership deal to power next generation of AI chips
- Microsoft's MAI-Image-2 launches for creators
Meta framed the support assistant as a practical utility rather than a conversational novelty, noting that users with account problems need resolution rather than suggestions.
Over the next few years, Meta plans to deploy its more advanced enforcement systems more broadly once they consistently outperform existing methods, reduce dependence on third-party vendors, and retain human reviewers for the highest-risk decisions.
The recap
- Meta launches Meta AI support assistant across Facebook and Instagram.
- Advanced AI systems found 5,000 scam attempts per day.
- Company plans to deploy systems across apps over the next few years.
