Meta AI launches global support assistant to resolve account issues on Facebook and Instagram

Meta AI Support Assistant Arrives Globally on Facebook and Instagram

Meta has announced the global rollout of its AI support assistant on Facebook and Instagram, marking a significant shift in how the company handles account issues for billions of users 1

1

2

2

. The Meta AI assistant is now available across Android, iOS, and desktop-based Help Center platforms, providing 24/7 assistance with response times typically under five seconds . This deployment aims to resolve user issues more efficiently than traditional methods that often involve clicking through vague Help Center articles or searching YouTube for solutions.

What sets this global AI support assistant apart is its ability to take direct action rather than simply providing suggestions. The assistant can handle password updates and profile settings, change post visibility, update user names, and configure various account settings 1

1

. These action-taking capabilities are currently limited to Facebook but will expand to Instagram in the future 2

2

. The assistant also supports reporting scams and impersonation accounts, managing privacy preferences, and appealing content removals 2

2

. For users in the US and Canada, the assistant can help with account access issues, a feature Meta plans to expand 1

1

.

AI-Powered Content Moderation Shows Promising Results

Beyond customer support, Meta AI is driving substantial improvements in content moderation and user safety across its platforms. The company's more advanced AI enforcement systems are catching approximately 5,000 scam attempts per day that previously went undetected by existing review teams 1

1

2

2

. Early testing has produced striking results: reports of celebrity impersonation have dropped by over 80%, while the systems identify twice as much violating adult content compared to human review teams 1

1

2

2

.

The AI systems have also reduced content enforcement mistakes by more than 60% 1

1

. These systems can detect subtle account takeovers and identify fake websites impersonating legitimate brands, with one test driving down views of scam ads by 7% 1

1

. The language coverage has expanded dramatically, now supporting languages spoken by 98% of internet users, up from approximately 80 languages previously 1

1

2

2

.

Human Oversight Remains Central to Content Enforcement

Despite the expanded role of AI in handling policy violations and scams, Meta emphasizes that human reviewers remain critical to its content enforcement strategy 1

1

2

2

. Human oversight will continue handling the most complex decisions, including account disablement appeals and law enforcement referrals 1

1

. The AI systems will focus on repetitive, high-volume tasks, freeing human reviewers to concentrate on nuanced decisions requiring human judgment 1

1

.

Meta plans to gradually deploy these AI systems over the coming years, suggesting a measured approach to automation 2

2

3

3

. The company stated it will introduce more advanced AI systems to improve content enforcement, helping detect and remove serious violations like scams and illegal content more accurately 3

3

. For users, this means faster resolution of account issues, improved protection from impersonation and scams, and a more responsive platform that can handle the scale of billions of users while maintaining the nuanced decision-making that complex situations demand.