MENLO PARK, July 12 — Meta’s artificial intelligence push is no longer a side experiment. It is the engine of the company’s future revenue, and the numbers coming out of its advertising business make the calculus plain. Meta launched its Meta AI assistant in September 2023, initially available on Facebook, Instagram, and WhatsApp in the United States.
The assistant, powered by Meta’s LLaMA 2 large language model, can generate text, answer questions, and create images. It marked the company’s first major consumer-facing AI product, but the broader stakes were always about monetization.
Strategically, Meta is betting that AI will do for its bottom line what social networking did for its growth. In December 2024, the company introduced AI-generated content tools for advertisers, allowing automated creation of ad copy and images. Early tests showed those tools improved click-through rates by an average of 32%.
That is not a marginal gain in a business where every percentage point of engagement shifts billions in ad spend.
The advertising fortress
Meta’s advertising revenue for 2024 was $127.5 billion, representing 94.5% of total revenue. The company is still overwhelmingly an advertising company, but what this signals is a shift in how that advertising gets made and targeted. AI-powered ad targeting increased average revenue per user by 8% in the fourth quarter of 2024.
For a platform with billions of users, that kind of lift is transformative. The company also launched AI-generated stickers and avatars for users in October 2024.
These features are not just entertainment. They are data collection and engagement tools that feed the AI models, improving the targeting algorithms that drive ad revenue.
Beyond advertising
Meta plans to introduce AI-generated video content in 2025. That move would put the company in direct competition with AI video startups and potentially open new revenue streams. These products are part of Meta’s strategy to monetize AI beyond advertising, including potential subscription tiers for advanced AI features.
The subscription model is still unconfirmed in detail, but the direction is clear. Meta wants to sell AI capabilities directly to users, not just to advertisers.
That would diversify a revenue base that is currently almost entirely dependent on ad sales. What this signals to the broader tech landscape is that Meta is all-in on AI as the next platform shift. The company is not waiting for the market to develop.
It is building the tools, training the models, and deploying the features now, with the expectation that the advertising returns will fund the long-term play. The question for competitors is whether they can match Meta’s scale advantage.
With $127.5 billion in annual ad revenue, Meta has the resources to invest in AI research and deployment at a level few companies can match. The broader stakes for the industry are clear: whoever controls the AI tools that generate and target content will control the next era of digital advertising.
