The Financial Times (FT) reported on the 24th local time that Meta, Facebook's parent company, is speeding up efforts to replace the work of reviewing whether content and ads posted on its key platforms violate rules with generative artificial intelligence (AI).
Meta announced this year that it would invest up to $145 billion (about 223 trillion won) in AI infrastructure and more, and it is seen as moving to cut expenses to offset the massive AI investment expense.
Citing sources familiar with the matter, FT said Meta is pushing to use AI to review content and ads on its platforms, which could yield expense savings of several billion dollars annually. Meta has already replaced half of its content review work with AI and plans to expand that share by the end of the year. For some content, it is also considering reducing the share of human review by more than 90%.
Meta said the reason for expanding AI-based content review is not expense savings but to use rapidly advancing technology more effectively. In initial tests conducted since March, the company found that in the process of policing rule-violating content, AI made on average 13% fewer mistakes than people. It also detected 10% more actual violations.
Meta said, "We only deploy advanced AI systems when they consistently deliver better performance than our existing content enforcement methods."
These AI replacement efforts come as Mark Zuckerberg, Meta chief executive officer (CEO), is pouring hundreds of billions of dollars into recruiting talent and building infrastructure to develop a "personal-level superintelligence," leading industry watchers to say the company is expanding AI adoption to cut expenses.
In fact, Meta is actively using AI to automate internal tasks such as coding to reduce operating expenses and has carried out multiple rounds of restructuring. Last month, Meta said it would lay off about 8,000 employees, roughly 10% of its global workforce.