Buy, buy, buy! Meta has set its sights on two AI video companies

Wallstreetcn
2025.08.01 04:22
portai
I'm PortAI, I can summarize articles.

Meta has listed two star startups in the AI video generation field—Pika and Higgsfield—as potential collaboration or acquisition targets. This move aims to enhance content generation capabilities for its social applications, smart glasses, and VR business, marking a key step in Zuckerberg's "personal superintelligence" blueprint. Previously, OpenAI and Google had released the video large models Sora and Veo, respectively

As the AI arms race intensifies, Meta is extending its mergers and acquisitions and collaboration efforts into the emerging field of AI video generation.

According to a recent report from The Information, insiders revealed that Meta has recently discussed potential collaboration with AI video startup Pika, with options including a direct acquisition or licensing Pika's technology. Additionally, Meta has also explored acquisition possibilities with Higgsfield, a video generation application focused on creators, but those negotiations are no longer ongoing.

In these discussions, Pika was founded in 2023 by two Stanford University computer science PhD dropouts, Demi Guo and Chenlin Meng, and has raised approximately $135 million from investors including Lightspeed Venture Partners. Higgsfield announced last April that it completed an $8 million seed round led by Menlo Ventures.

For Zuckerberg, deals with AI video companies are a significant boost to realizing his vision of "personal superintelligence." In media interviews, he stated that this vision "focuses on entertainment, culture, and personal relationships with people." AI technology that can generate and understand video not only has the potential to greatly enrich the content ecosystem of Meta's social applications but also provides critical technological support for its future core smart glasses and the virtual reality (VR) business, which has already invested billions of dollars, such as real-time generation of virtual scenes.

Last month, Meta introduced AI video editing features in its AI assistant, and Zuckerberg mentioned during the earnings call that these features "have seen early progress," adding that "the content itself has great potential to improve" and "there is still a lot of work to be done here."

Meta is not starting from scratch in the field of video generation; its video editing capabilities are built on prior research, including the Movie Gen model showcased last October. However, facing rapid iterations from competitors, Meta is clearly feeling the pressure. OpenAI's Sora and Google's Veo models have demonstrated astonishing capabilities in video generation quality and realism, attracting significant industry attention.

Meta has long shown interest in AI video generation technology. In addition to recent contacts with Pika and Higgsfield, sources indicate that Meta also discussed collaboration, including acquisition, with another leading AI video company, Runway, earlier this year, but those conversations did not continue.

A Broader AI Strategic Restructuring

These acquisition intentions are not isolated events but are a key part of Meta's broader AI strategic restructuring. Just last month, the company announced the appointment of Alexandr Wang, CEO of Scale AI, as its Chief AI Officer and agreed to invest $14.3 billion in the data labeling company.

At the same time, former GitHub CEO Nat Friedman and former Safe Superintelligence CEO Daniel Gross have also joined Meta to co-lead the newly established AI team—Meta Superintelligence Lab In order to recruit for its new laboratory, Meta has poached dozens of researchers from competitors such as OpenAI, Anthropic, and Google in recent weeks, and acquired the voice AI startup PlayAI to strengthen its talent pool