Meta's open AI model family, Llama, has reached a significant milestone, surpassing 1 billion downloads, according to CEO Mark Zuckerberg. The announcement, made on Threads, highlights a rapid rise in adoption, with downloads increasing by 53% since December 2024. Llama powers Meta's AI assistant across Facebook, Instagram, and WhatsApp, forming a crucial part of the company's expanding AI ecosystem.

Despite its success, Llama has not been without controversy. Meta faces a lawsuit alleging the model was trained on copyrighted material without permission, while regulatory concerns have stalled its rollout in some European markets. Additionally, emerging competitors, such as China's DeepSeek R1, have challenged Llama's technological edge, prompting Meta to intensify its AI research efforts.

Looking ahead, Meta plans to launch several new Llama models, including those with advanced reasoning and multimodal capabilities. Zuckerberg has hinted at 'agentic' features, suggesting the AI could soon perform tasks autonomously. More details are expected at LlamaCon, Meta's first AI developer conference, set for 29 April.