Meta is actively collaborating with the public sector to integrate Llama across various segments of the U.S. government, as announced by CEO Mark Zuckerberg. This statement was made during the company’s Q3 earnings call. The announcement has led to several questions, such as which government departments will implement Meta’s AI models, the intended applications of these models, possible military-specific uses, and whether Meta is receiving compensation for these collaborations.
Similar collaborations with the government are being explored by Meta’s AI competitors. OpenAI and Anthropic have indicated plans to share their models with the U.S. AI Safety Institute for pre-emptive safety evaluations. Google’s intermittent relationship as an AI vendor for the Pentagon is also well-documented. Furthermore, OpenAI, in a recent blog post, stated that its models are utilized by DARPA, the U.S. Agency for International Development, and the Los Alamos National Laboratory.
In anticipation of further details on Meta’s AI involvement with the government, Zuckerberg provided insights into the upcoming Llama model during the earnings call. He noted that version four is currently being trained on an unprecedentedly large cluster, with expectations for “new modalities,” enhanced reasoning abilities, and significantly faster performance upon its release next year.
Zuckerberg also mentioned Meta’s plans to increase AI investments in 2025, acknowledging that this might not align with short-term investor expectations, but he foresees significant long-term benefits. He expressed enthusiasm for the ongoing developments, citing it as potentially the most dynamic period he has witnessed in the industry, and emphasized the importance of capitalizing on emerging opportunities.
Financially, Meta continues to exhibit growth, reporting a revenue of $40.5 billion for Q3, reflecting a 19 percent increase from the previous year, with profits amounting to $17.3 billion. The daily usage of its apps has reached 3.29 billion people, marking a 5 percent growth from the prior year.