MW Is Meta's pricey AI hiring spree worth it? This analyst has doubts.
By Britney Nguyen
Meta's push for top-tier AI talent, which could cost up to $1.4 billion, raises questions about its approach to competing with other companies, a Rosenblatt analyst says
Meta Platforms Inc.'s hiring spree for top-tier artificial-intelligence talent has one analyst questioning the company's approach to competing in the AI race and wondering whether the moves bring near-term financial risk.
Late last month, Meta $(META)$ Chief Executive Mark Zuckerberg announced the Meta Superintelligence Labs, which he said would focus on developing advanced AI that can handle tasks at the same level as, or better than, humans. Earlier in the month, it was reported that Zuckerberg offered some talent, including those poached from OpenAI, $100 million pay packages to work on the team. The tech giant also invested $14 billion in data-labeling AI startup Scale AI and hired its founder and former CEO Alexandr Wang to head the new AI effort.
See more: Meta's stock hits a record high as Mark Zuckerberg goes on an AI hiring spree
The MSL team includes 14 new hires, which Rosenblatt Securities analyst Barton Crockett said would cost $1.4 billion in the first year if all were paid $100 million. However, he noted that one new hire, Lucas Beyer, said on X that he and two of his poached colleagues "did not get 100M sign-on," calling rumors around that big number "fake news." Crockett also cited Meta Chief Technology Officer Andrew Bosworth reportedly telling employees that the packages are for a few people in senior leadership roles and not a signing bonus, according to The Verge.
In the first quarter of this year, Meta lowered its expense guidance for 2025 by $1 billion to between $113 billion and $118 billion, Crockett said. He added that Meta could've known "of this MSL plan then."
"It is possible that Meta is seeing enough additional efficiency from AI boosts to programmer productivity/staffing reductions to more than fund this MSL push," Crockett said.
Even if Meta can afford it, Crockett said "it seems probably not great for the giants of [large language models] to bleed off their capital into the mutual-assured carnage of endless battles for talent." It's possible, he said, that the company "sees its efforts here as more one and done than ongoing."
Meta declined to comment for this story.
Crockett also questioned Meta's strategy to compete with other AI model makers for leadership in AI benchmarks with its Llama models, adding that the company is not a competitor for AI search engines and chatbots.
If Meta is trying to compete with Alphabet Inc. $(GOOGL)$'s Google for AI-powered search, for example, "then investing in the best LLM might make sense," Crockett said. He added that the Rosenblatt team is doubtful that Meta wants to compete as a cloud service provider.
"In any event, for now AI is probably most valuable to Meta in specialized capacities of improving advertiser productivity, and enhancing social media experiences for users," Crockett said. "It's not clear to us that Meta wins a lot, beyond bragging rights, in shooting for the best LLM benchmark scores, unless it takes on Google in AI search."
Meanwhile, Crockett said, "margin compression is a growing overhang risk" for Meta, which saw its first-quarter sales grow 16%. The company's second-quarter guidance calls for sales growth of up to 16%, Crockett said, and growth for the rest of the year could come in at a midteens rate. But expenses are expected to grow faster than that, he noted, based on the company's full-year forecast. Meta will report its second-quarter results on July 30.
The new MSL hires "might in part reflect confidence in revenue trends, as well as expense efficiency," Crockett said. But if the company's margins "start eroding," Crockett said it would likely affect how investors feel about Meta's stock.
-Britney Nguyen
This content was created by MarketWatch, which is operated by Dow Jones & Co. MarketWatch is published independently from Dow Jones Newswires and The Wall Street Journal.
(END) Dow Jones Newswires
July 07, 2025 11:16 ET (15:16 GMT)
Copyright (c) 2025 Dow Jones & Company, Inc.