Mergers and acquisitions+heavy investment Meta (META.US) takes a multi-pronged approach to sprint to the AI circuit

The Zhitong Finance App learned that Meta Platforms (META.US) is reportedly negotiating cooperation with artificial intelligence (AI) video company Pika and discussing potential acquisitions or technology licensing deals. The report added that the social and tech giant also had acquisition negotiations with Higgsfield, an image and video generation application company mainly aimed at creator users, but the negotiations have now stopped.

According to reports, since the beginning of this year, Meta has been actively engaged in AI-related transactions. In June of this year, Meta acquired 49% of the shares in the data platform Scale AI for nearly $15 billion, then appointed Alexandr Wang, CEO of Scale, who is only 28 years old, as Meta's AI leader. Meta also acquired PlayAI, a small AI voice generation startup in July, which focuses on anthropomorphic voice generation, but the terms of the deal were not disclosed. Prior to that, Meta was also trying to acquire Perplexity AI, Runway AI, and FuriosaAI.

At the same time, Meta is also gathering high-end AI talents from competitors such as Google (GOOGL.US), OpenAI, and Apple (AAPL.US) and forming a strong “Meta Superintelligence Labs (MSL)” team by providing huge salaries. In July alone, Meta brought in a number of researchers from OpenAI, including Shengjia Zhao, Jiahui Yu, Shuchao Bi, Hongyu Ren, Lucas Beyer, Alexander Kolesnikov, and Xiaohua Zhai. Meta also recruited Daniel Gross, CEO of startup Safe Superintelligence, and Nat Friedman, former CEO of GitHub, to join MSL. Earlier this month, Meta successfully removed Ruoming Pang, the head of Apple's large language model team, with a $200 million salary package, and then took Mark Lee and Tom Gunter from Apple. The two had previously worked under Ruoming Pang.

To support the advancement of its super-smart program, Meta is seeking financing of $29 billion, of which $3 billion comes from private equity companies such as Apollo Global Management (APO.US), KKR (KKR.US), Brookfield (BAM.US), Carlyle (CG.US), and Pimco, while the remaining $26 billion will be raised through debt.

It's worth mentioning that Meta announced strong second-quarter revenue and third-quarter guidance this week, indicating that the tech giant's advertising business is still growing rapidly, enough to support its huge AI spending. According to the data, Meta's second-quarter revenue was US$47.52 billion, up 22% year on year, better than market expectations; after dilution, earnings per share were 7.14 US dollars, up 38% year on year, which was also better than market expectations. Meta expects third-quarter revenue to be between $47.5 billion and $50.5 billion, and the median value in this range exceeded analysts' average expectations of $46.2 billion.

Additionally, Meta raised the lower limit of the 2025 capital expenditure forecast range as the company continues to invest heavily in talent, infrastructure, data centers, and energy to remain competitive in the rapidly evolving AI competition. The company currently anticipates spending between $66 billion and $72 billion this year. In view of ongoing trade disputes and AI investments, this forecast was already raised in April. Meta said that costs will continue to grow at a faster rate next year, especially when spending on infrastructure construction and employee compensation will increase further.

Matt Britzman, stock analyst at Hargreaves Lansdown, said, “All of these expenses will have some impact on the company's profitability in the short term, but in the long run, Meta is expected to be a clear winner in the field of artificial intelligence.” He said Meta performed “extremely well” in the second quarter, in part because improvements in AI enabled the company to raise the average price of ads.