Meta CEO Mark Zuckerberg previously stated that “selling access” to Llama AI models was not part of the company’s business model. However, according to newly unredacted court documents in the copyright lawsuit Kadrey v. Meta, it’s been revealed that Meta profits from Llama AI through revenue-sharing agreements with companies that host its models. The filing confirms that Meta receives a share of the revenue generated by host companies, although specific hosts are not identified. Partners such as AWS, Nvidia, Databricks, Groq, Dell, Azure, Google Cloud, and Snowflake have been identified in various Meta blog posts as Llama hosts.
Llama AI’s Integration and Monetization
While developers can download and run Llama models independently on their hardware, many opt to use hosts for additional services that make deployment easier. The disclosure of these revenue-sharing agreements contrasts with Zuckerberg’s earlier statements about Meta’s business strategy, indicating a more direct path to profit through collaborations with companies like Microsoft, Amazon, and Google, which resell Llama-powered services.
Despite Meta’s plans to scale its AI investments, the legal challenges surrounding Kadrey v. Meta are serious. Plaintiffs accuse Meta of using pirated e-books for training Llama, and they allege that the company facilitated infringement through torrenting methods, distributing pirated content in the process. As Meta defends itself in court, its ongoing AI strategies are at risk of facing serious consequences.
In addition to addressing these legal matters, Meta plans to invest significantly in AI research and infrastructure, with plans to allocate $60 billion-$80 billion in 2025. To offset some of these investments, the company is reportedly considering a subscription service for Meta AI, which would bring new capabilities to its AI assistant.
What The Author Thinks
While Meta’s business strategies in AI showcase a robust model for technological advancement and profitability, they also pose serious questions about the ethical boundaries of data usage and copyright adherence. The ongoing lawsuit and its revelations could potentially reshape how tech companies approach AI training methodologies and their legal ramifications. Meta’s handling of this situation will be crucial in setting precedents for AI development and copyright laws in the tech industry.
