The company is aggressively hiring top AI talent, offering lucrative compensation packages that often exceed $1 million, to bolster its AI capabilities.

Meta is investing heavily in infrastructure, planning to deploy 600,000 GPUs by the end of 2025 to build one of the largest AI training capabilities globally.

This heavy investment in AI, including new data centers and high salaries for researchers, is part of Meta’s broader effort to develop artificial general intelligence (AGI) and diversify revenue sources beyond advertising.

CEO Mark Zuckerberg is actively involved in these initiatives, emphasizing the importance of balancing innovation with ethical practices and transparency.

Meta is considering a major strategic shift in its AI approach, potentially moving from an open-source model to a closed model, a decision that could reshape its position in the AI landscape.

Despite these strategic discussions, Meta is expected to continue its revenue growth, with Q2 2025 guidance estimating revenues between $42.5 billion and $45.5 billion.

However, the company faces challenges including a $4.2 billion loss in Reality Labs, regulatory hurdles in the EU, and increased capital expenditures for AI infrastructure.

Meanwhile, Meta is experiencing significant growth in user engagement, especially in AI-driven advertising across platforms like Facebook, Instagram, and WhatsApp.

Some praise Meta’s transparency and innovative approach to AI, even as it faces criticism and regulatory scrutiny.

The potential move to a closed model could allow Meta to explore new monetization opportunities but might also cause it to lose its leadership in open-source AI, especially against Chinese firms like DeepSeek and Moonshot AI.

Meta’s most powerful AI model, Behemoth, was completed but its release has been postponed due to poor internal performance, and testing has reportedly stopped since the lab's launch.

Any final decision on shifting strategies will require approval from CEO Mark Zuckerberg, with top lab members, including new chief AI officer Alexandr Wang, involved in the discussions.