Meta Platforms Reconsiders AI Strategy, Weighs Use of Rival Models Over Llama
Meta’s Shift Highlights Performance Concerns and Intensifies AI Talent Race
Meta Platforms (META) is reportedly evaluating a move away from its own Llama language models in favor of adopting AI models from leading competitors like OpenAI and Anthropic, as reported by The New York Times.
- This potential pivot marks a significant shift, since Llama models are open-source, allowing broad access and modification, while OpenAI and Anthropic operate with closed, proprietary systems.
No final decision has been made, but the reconsideration follows the underwhelming launch of Meta’s Llama 4 models in April.
- The models were introduced during the company’s LlamaCon event, but many developers were disappointed by their performance compared to rivals.
- Meta has since increased spending to close the technology gap, including a $14.3 billion investment in Scale AI, which resulted in Scale’s CEO Alexandr Wang joining Meta to head its “superintelligence” team for artificial general intelligence.
Aggressive Moves to Attract Top AI Talent
Meta’s recent strategies include offering signing bonuses as high as $100 million to attract researchers from OpenAI.
- The company is also in talks to acquire PlayAI, a startup specializing in realistic AI voices, and has previously considered acquiring Perplexity AI, Runway AI, and FuriosaAI before opting for the Scale AI deal.
- These actions underline Meta’s commitment to strengthening its position in the AI field under CEO Mark Zuckerberg.
Analyst Outlook for Meta Stock
On Wall Street, analysts maintain a Strong Buy consensus rating for META stock.
- The average price target is $716.48 per share, suggesting that the stock is trading near fair value based on current forecasts.