In a message to employees on Wednesday, Alexandr Wang, Meta’s chief artificial intelligence officer, said the company would be laying off 600 people in its A.I. division. The cuts, he said, would help Meta build new products faster.
“By reducing the size of our team, fewer conversations will be required to make a decision,” Mr. Wang wrote in an internal memo viewed by The New York Times.
But buried amid the A.I. division layoffs were a different set of cuts. The company laid off more than 100 people in its risk review organization, according to three people familiar with the move and internal memos viewed by The Times. That group is largely staffed by employees responsible for making sure Meta’s products abide by an agreement with the Federal Trade Commission as well as privacy rules set by regulatory bodies around the world, the people said.
In a note sent to employees on Wednesday, Michel Protti, Meta’s chief privacy officer, said the company would be shrinking the risk team and replacing most of its manual reviews with an automated systems.
“By moving from bespoke, manual reviews to a more consistent and automated process, we’ve been able to deliver more accurate and reliable compliance outcomes across Meta,” Mr. Protti said in the memo. “We remain committed to delivering innovative products while meeting our regulatory obligations.”
Mr. Protti did not specify how many roles were being cut. But insiders described the cuts as a “gutting” of the workers in the department who review projects at Meta for privacy and integrity risks, according to two employees familiar with them. Meta is laying off many of the workers on the risk review team in the London office and probably more than 100 people in the risk organization across the entire company.
“We routinely make organizational changes and are restructuring our team to reflect the maturity of our program and innovate faster while maintaining high compliance standards,” a Meta spokesman said in a statement. Some details of the cuts were earlier reported by Business Insider.
The moves come amid a broader overhaul of Meta’s organizational structure. Mark Zuckerberg, Meta’s chief executive, has spent the past three years shaking up his company as it races to compete with new rivals like OpenAI, the maker of the ChatGPT chatbot.
But Meta executives have become frustrated with the pace of product development, according to three people who spoke to The Times. One division holding things up — by design — was the company’s risk organization.
In 2019, the F.T.C. required Meta, then known as Facebook, to add new positions and practices to increase the transparency and accountability of how it treats user information. The agency also imposed a record $5 billion fine against Facebook for deceiving users about their ability to control the privacy of their personal data.
The risk organization is responsible for overseeing and auditing all new products for potential threats to user privacy or changes that could violate the F.T.C. consent order the company agreed to in 2019.
In 2020, Mr. Protti said the changes would bring about “a new level of accountability” and made sure that privacy was “everyone’s responsibility at Facebook.”
Current and former employees in the risk organization said they were skeptical that replacing them with automated systems would be as effective, particularly around issues as sensitive as user privacy. Meta has spent the better part of the past decade being closely monitored by the F.T.C. and the Justice Department in the United States while facing intense scrutiny from regulatory bodies in Europe.
Over the past year, Meta has slowly introduced automation into its risk auditing process by dividing potential issues into two parts: “Low risk” updates to new products were subject to automated review and later audited by humans. “High or novel risk” product issues were subject to the more intense process of immediate review by human auditors, according to two people familiar with the process.
In August, Meta split its artificial intelligence division into four parts: FAIR, the company’s research division; TBD Labs, which works on so-called superintelligence in A.I. systems; one division focused on building new products; and a fourth organization called infrastructure, such as data centers and A.I. hardware.
Beyond Meta’s risk organization, other cuts on Wednesday targeted veteran members of Meta’s FAIR team and those who had worked on previous versions of Meta’s open source A.I. models, called Llama. Among the employees who were laid off was Yuandong Tian, FAIR’s research director, who had been at the company for eight years.
But there was one division that was spared: TBD Labs, the organization largely made up of new, highly paid recruits working on the next generation of A.I. research. The department is led by Mr. Wang.
Mike Isaac is The Times’s Silicon Valley correspondent, based in San Francisco. He covers the world’s most consequential tech companies, and how they shape culture both online and offline.
Eli Tan covers the technology industry for The Times from San Francisco.
The post Meta Layoffs Included Employees Who Monitored Risks to User Privacy appeared first on New York Times.
