Newly unsealed court documents show the Meta generative AI revenue forecast could hit as high as $1.4 trillion by 2035, signaling massive long-term ambitions. The internal projection also expects Meta’s generative AI tools to generate up to $3 billion in revenue by 2025.
The filings, presented in a copyright infringement case by authors claiming Meta misused their content to train AI, stop short of defining what qualifies as a “generative AI product.” Still, Meta’s monetization strategy for AI is already visible in its partnerships and product ecosystem.
AI Products, APIs, and Potential Monetization Paths
Meta currently profits from multiple avenues tied to generative AI. These include revenue-sharing deals with platforms that distribute its open-source Llama AI models, and the recent release of a cloud-based API designed to help developers fine-tune and evaluate Llama.
During Meta’s Q1 earnings call, CEO Mark Zuckerberg indicated that Meta AI, the company’s assistant, might eventually adopt monetization models like ads and premium subscription features, adding another revenue stream.
Meta’s Billion-Dollar AI Investment Strategy
The same documents show Meta’s financial commitment to generative AI is already enormous. In 2024, the company spent more than $900 million on GenAI initiatives, with projections that the figure could surpass $1 billion in 2025. This budget excludes capital required to scale the infrastructure needed to train and deploy large AI models.
Meta has publicly stated it intends to allocate $60–$80 billion in capital expenditures in 2025 alone, largely for constructing AI-ready data centers to handle growing computational needs.
Licensing vs. Piracy: Disputed Content Sources
The filings also reveal that Meta considered allocating as much as $200 million for AI training data in 2023, with nearly $100 million earmarked specifically for book licenses. However, the company allegedly opted for alternative — and controversial — methods: sourcing copyrighted content without authorization, according to the lawsuit.
This claim, if true, could have legal ramifications as the debate around fair use in AI training heats up globally.
Meta Responds to Copyright Allegations
In a statement to TechCrunch, a Meta spokesperson defended the company’s use of training materials:
“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.”
Meta’s AI Vision at Scale
The Meta generative AI revenue forecast projects earnings between $460 billion and $1.4 trillion by 2035, with an estimated $2 to $3 billion in revenue expected in 2025. To support this vision, Meta invested over $900 million in GenAI initiatives in 2024, with forecasts suggesting spending could surpass $1 billion in 2025. Additionally, the company plans to allocate $60–$80 billion toward infrastructure development, primarily for new AI-ready data centers.
On the legal front, Meta is facing a lawsuit over allegations of unauthorized use of books for AI training, with court documents revealing it once considered spending up to $200 million on licensed training data, including $100 million specifically for books. Instead, the company allegedly pursued more controversial alternatives.
To monetize its generative AI assets, Meta is focusing on licensing its Llama models, offering cloud-based APIs, and potentially integrating ads and subscription features into its Meta AI assistant.
Get the Latest AI News on AI Content Minds Blog