Meta forecasted it would make $1.4T in revenue from generative AI by 2035

Last year, Meta boldly predicted that its generative AI products would generate between $2 billion and $3 billion in revenue by 2025. Looking further ahead, they anticipated a staggering growth to between $460 billion and $1.4 trillion by 2035. These projections were revealed in court documents that surfaced on Wednesday, amidst a lawsuit by book authors against Meta. They allege the company trained its AI on their works without permission. However, the documents don't specify what Meta classifies as a "generative AI product." It's no secret, though, that Meta has been tapping into various forms of generative AI to boost its earnings and plans to continue doing so.

Meta has lucrative revenue-sharing agreements in place with companies that utilize its open-source Llama models. Recently, they launched an API that allows users to customize and evaluate these models. Additionally, Meta AI, the company's AI assistant, could soon start displaying ads and offer a premium subscription with enhanced features, according to CEO Mark Zuckerberg during the Q1 earnings call on Wednesday.

Meta's Significant Investment in AI

The court documents also shed light on Meta's hefty investments in its AI product groups. In 2024, Meta's "GenAI" budget reportedly exceeded $900 million, and it's projected to surpass $1 billion this year. This doesn't even account for the costs of the infrastructure required to operate and train these AI models. Meta has also earmarked between $60 billion and $80 billion for capital expenditures in 2025, primarily to build expansive new data centers.

These budgets could have been even higher if Meta had pursued deals to license books from the authors suing them. The documents reveal that in 2023, Meta considered spending up to $200 million to acquire training data for Llama, with approximately half of that earmarked for books. However, the company allegedly opted for a different route—massively pirating ebooks.

In response, a Meta spokesperson provided the following statement to TechCrunch:

“Meta has developed transformational [open] AI models that are powering incredible innovation, productivity, and creativity for individuals and companies. Fair use of copyrighted materials is vital to this. We disagree with [the authors’] assertions, and the full record tells a different story. We will continue to vigorously defend ourselves and to protect the development of generative AI for the benefit of all.”

Exhibit at TechCrunch Sessions: AI

Secure your spot at TC Sessions: AI and show 1,200+ decision-makers what you’ve built — without the big spend. Available through May 9 or while tables last.

Looking Ahead

As Meta continues to push the boundaries of generative AI, the financial stakes are high, and so are the legal challenges. The company's commitment to defending its practices and advancing AI technology remains unwavering, even as it faces scrutiny over how it sources its training data.