Meta's Generative AI Revenue Forecast: $1.4 Trillion by 2035

Meta's Projected Revenue from Generative AI

Last year, Meta forecasted that its generative AI products would generate between $2 billion and $3 billion in revenue by 2025. Further projections, detailed in recently unsealed court documents, estimate potential revenue ranging from $460 billion to $1.4 trillion by 2035.

AI Training Lawsuit Details

These figures were presented in documents submitted by legal representatives of authors who are suing Meta. The authors allege unauthorized use of their copyrighted works in the training of Meta’s AI models.

The precise definition of a “generative AI product” as used by Meta remains unspecified within the documents. However, it is widely known that the technology company currently profits, and anticipates increased profits, from various applications of generative AI.

Revenue Streams and Model Access

Meta currently has revenue-sharing agreements with companies hosting its open Llama model collection. Additionally, the company has recently introduced an API designed for the customization and evaluation of Llama models.

Meta AI, the company’s AI assistant, is also expected to generate revenue through advertising and potentially a subscription service offering enhanced features, as stated by CEO Mark Zuckerberg during the Q1 earnings call.

Significant Investment in AI Development

The court filings also reveal substantial financial investment by Meta in its AI product groups.

In 2024, the company’s “GenAI” budget exceeded $900 million. Projections indicate this figure could surpass $1 billion in the current year.

This budget does not include the costs associated with the infrastructure required for running and training AI models. Meta has previously announced plans to invest between $60 billion and $80 billion in capital expenditures for 2025, primarily focused on expanding its data center capacity.

Potential Licensing Costs and Data Acquisition

These budgetary allocations could have been even larger if Meta had pursued licensing agreements for books from the authors involved in the lawsuit. Documents indicate that in 2023, Meta considered spending over $200 million to acquire training data for Llama.

Approximately $100 million of this amount was earmarked for books. However, the company reportedly opted for alternative methods, allegedly involving large-scale unauthorized ebook acquisition.

Meta's Official Response

A Meta spokesperson provided TechCrunch with the following statement:

“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.”