The Menlo Park-based social media giant says its new artificial intelligence enforcement technology outperforms human review teams on key metrics, such as detecting fake accounts and sexual solicitation content.<\/p>\n
The company announced on its website Thursday that it is \u201crolling out the Meta AI support assistant globally on Facebook and Instagram, not just to provide 24\/7 help for account issues like updating passwords and profile settings, but also \u201cto transform our approach to content enforcement, more accurately finding and removing severe content violations like scams and illegal content.\u201d<\/p>\n
Meta said the AI expansion will occur \u201cover the next few years.\u201d<\/p>\n
\u201cToday we\u2019re launching new AI tools for support and content enforcement on our apps to make them work better for you,\u201d the company said. \u201cAs technology advances, we\u2019re applying AI in more ways so you can get reliable, action-oriented help when you need it, and we can catch more severe violations like scams faster and more accurately, with fewer over-enforcement mistakes.\u201d<\/p>\n
Launching the Meta AI Support Assistant<\/p>\n
In December, Meta previewed its AI support assistant and it\u2019s now rolling it out in countries and territories where Meta AI is currently available on the Facebook and Instagram apps for iOS and Android, and within Help Center on Facebook and Instagram.<\/p>\n
The new Meta AI support assistant is designed to help resolve account problems and offers answers for\u00a0 question about notification settings or new features and also can handle:<\/p>\n
Reports of scams, impersonation accounts, or problematic content
\nQuestions about why content was taken down and how to appeal these decisions
\nManaging\u00a0 privacy settings
\nResetting passwords
\nUpdating profile settings<\/p>\n
The Meta AI support assistant is built into Facebook and Instagram, and Meta promised responses \u201ctypically in under five seconds.\u201d<\/p>\n
Meta called its AI support assistant \u201ca major step in our work to deliver stronger support on our apps.\u201d<\/p>\n
It is being rolled out in all languages supported by Facebook and Instagram for support topics.<\/p>\n
Improving Content Enforcement<\/p>\n
Faced with increasing criticism about its apparent easing of content moderation, Meta said it is continuing working on \u201ccutting down on mistakes and focusing our proactive enforcement toward illegal and the most severe content on our platforms like terrorism, child exploitation, drugs, fraud, and scams.\u201d<\/p>\n
Meta said it is experimenting with more advanced AI systems for content enforcement that \u201cwe believe can catch more of these violations more accurately while also stopping more scams and responding faster to real-world events with fewer over-enforcement mistakes.\u201d<\/p>\n
Meta said its new AI systems can:<\/p>\n
Reduce the chance that scammers trick people into giving away their login details, ultimately finding and mitigating 5,000 scam attempts per day that no existing review team had caught before
\nIdentify and prevent more accounts from impersonating celebrities and other high-profile people, which helped us to reduce user reports of the most impersonated celebrities by over 80%
\nCatch two times more violating adult sexual solicitation content than our review teams, while also decreasing the rate of mistakes by more than 60%
\nPrevent an account takeover by noticing it was suddenly accessed from a new location, the password was changed, and edits were made to the profile \u2014 changes that, in isolation, look harmless to a person reviewing the account, but AI was able to recognize as a threat
\nDetect a fake site spoofing a legitimate web address and pretending to be a popular sporting goods store by noticing the real logo being used with unusually low prices and a suspicious web address.<\/p>\n
\u201cThese more advanced AI systems can do all of this in languages spoken by 98% of people online \u2014 far beyond our previous coverage of around 80 languages,\u201d according to Meta.<\/p>\n
More advanced AI systems<\/p>\n
Over the next few years, we\u2019ll be deploying these more advanced AI systems across our apps once we\u2019ve seen them consistently perform better than our current methods of content enforcement, transforming our approach.<\/p>\n
Meta says it will reduce its reliance on third-party vendors for content enforcement and focus on strengthening internal systems and workforce.<\/p>\n
\u201cWhile we\u2019ll still have people who review content, these systems will be able to take on work that\u2019s better-suited to technology, like repetitive reviews of graphic content or areas where adversarial actors are constantly changing their tactics, such as with illicit drugs sales or scams.\u201d<\/p>\n
\u201cAI can help us move faster and operate at scale, but it doesn\u2019t replace human judgment \u2014 it helps us apply it more consistently across billions of pieces of content on our platforms,\u201d the company said in its announcement. \u201cExperts will design, train, oversee, and evaluate our AI systems, measuring performance and making the most complex, high\u2011impact decisions. For example, people will continue to play a key role in how we make the highest risk and most critical decisions, such as appeals of account disablement or reports to law enforcement.\u201d<\/p>\n
Meta also pledged that its Community Standards won\u2019t be changed as part of the shift to AI, and that it will be \u201c improving our methods for reporting violating content and for appealing mistakes.\u201d<\/p>\n
\u00a0<\/p>\n