Despite ongoing staffing shortages in numerous key areas, layoffs have been a running theme at major tech firms for some time. The latest news in this area comes from Meta, which announced a new planned layoff of personnel in the company’s risk division; the difference here is that it appears many of these positions will be replaced by AI tools. This includes numerous staff responsible for privacy compliance and AI safety.
The announcement follows a very recent layoff of 600 employees from its “Superintelligence Labs” that heads up its AI research & development. Chief compliance and privacy officer of product Michel Protti told the risk division workers that the layoffs were mostly driven by the company shifting from manual to automatic review processes.
Meta says it “doesn’t need” as much privacy compliance staff as it once did due to AI
An internal memo sent to risk division workers on October 29 by Protti outlined the company’s intention to lay off more privacy compliance staff, as part of cost-cutting measures that shift more duties from humans to AI.
Protti’s memo touted the company’s AI adoption as making privacy compliance review processes more consistent, accurate and reliable, removing the need for many “routine” decisions. This was framed as “freeing our teams to focus on the most complex and high-impact challenges,” though clearly this will not apply to the staff being let go.
The memo indicated that privacy compliance cuts are specifically coming to the Product Risk Program Manager, Shared Services, and Global Security & Privacy (GSP) teams. The GSP team is to be merged with the data protection officer’s team and the Readiness division, forming a new entity called the Regulatory Compliance Programs. The memo also indicated Areas jobs in London were marked for consolidation.
The exact number of cuts to expect from this round of layoffs remains unknown, but the announcement follows 600 layoffs from the central AI R&D team one week prior. That move was also revealed by reporting on an internal memo, issued by chief AI officer Alexandr Wang, who indicated that each remaining member would be expected to be more “load-bearing” and that the move would “speed up conversations” internally about product decisions. The company said it remains committed to attracting top AI development talent, however, following some reports of record-breaking salary offers in recent months. Several internal sources told the New York Times that at least 100 privacy compliance staff were being either let go or moved to other jobs.
A June LinkedIn post from Meta Privacy Compliance & AI Policy Leader Rob Sherman indicates that some of the layoffs may be tied to the recent development of an internal automated tool that automatically flags legal and policy requirements that apply to specific products. Sherman said that this tool is not involved in making decisions about risk and is primarily meant to reduce the time experts need to spend on these decisions instead.
AI job replacements come for AI development staff
While the alarm has been sounded about mass loss of employment due to AI for years now, AI workers are sometimes finding themselves among the early waves of casualties. Some other departments, particularly customer service, are also seeing disproportionate early job losses as more advanced “agentic AI” designed specifically for workforce replacement begins to roll out.
The privacy compliance department is hardly the first at Meta to see cuts due to some of the company’s own work on AI. New applicants to the tech giant are already seeing AI in parts of its interview and testing process that used to be administered by people. And more will be coming in the future, as CEO Mark Zuckerberg expects that agentic AI will be doing the work of “mid-level engineers” within a matter of months.
This trend is slowly but surely catching on everywhere, elsewhere in the tech industry and also well beyond. In July, Microsoft estimated that it terminated some 15,000 positions due to adoption of AI tools. Google has estimated laying off a total of 12,000 employees in recent years to free up resources for AI development. Salesforce CEO Marc Benioff recently announced that its own Agentforce software had become so capable that 4,000 job cuts in customer support can be expected in the near future. The big names in the finance industry are also working on similar agentic AI replacements for similar job roles and consequently slowing down hiring.
Meta has to proceed somewhat carefully with its risk and privacy compliance department, however, as it is in something of a unique position. It was ordered to create the current form of its risk organization in 2019 as part of a user data privacy settlement with the FTC, which also came packed with a record $5 billion fine. Certain positions and practices are compulsory under this agreement, which was the result of violation of a prior 2012 FTC order, and Meta remains bound by its terms until 2039.
