Meta Platforms CEO Mark Zuckerberg stated on Wednesday that the tech giant's enterprise artificial intelligence (AI) opportunities extend significantly beyond the AI agents introduced in June to assist businesses with customer service and daily operations.
Meta's Expansive Enterprise AI Vision
During the company's second-quarter earnings call, Zuckerberg elaborated on Meta's strategic expansion into the enterprise sector. He indicated that the company sees a substantial market for selling APIs, business agents, compute resources, and other services to large corporate clients.
These new revenue streams could diversify Meta's income beyond its primary advertising business and its growing subscription services. The company's initial focus within this enterprise AI push will be on enhancing its existing advertiser base by providing AI agents capable of operating across messaging platforms and other digital touchpoints.
“And, just like the ad system, effectively, we will get paid when we deliver results for those businesses,” Zuckerberg said, as quoted in the company's earnings call. He characterized this as an extension of existing sales and partnership efforts with millions of advertisers and hundreds of millions of small businesses utilizing Meta's platforms.
Expanding Beyond Small Business Clients
Zuckerberg also detailed plans for Meta to leverage its internal development tools for external enterprise customers. The company is currently building coding, development, and internal productivity tools for its own needs, but sees a significant opportunity to offer these to both small and larger businesses.
He acknowledged that engaging with enterprise clients represents a "different muscle" for Meta compared to its historical focus. However, the company appears poised to adapt to this new market dynamic.
Compute Sales and Future Infrastructure
Regarding the potential sale of compute resources to enterprise customers, Meta is working to balance immediate revenue needs with its long-term strategic plans. The company indicated it currently has opportunities to sell compute at a "significant premium over what we paid for it."
However, Zuckerberg cautioned investors against liquidating all compute resources for short-term gains, describing Meta's approach as a "portfolio" strategy. This approach integrates both long-term and short-term objectives for its compute infrastructure, particularly as the company moves toward what he termed "personal superintelligence," which will require advanced hardware for seamless interaction.
Broader AI Ambitions Beyond Enterprise
Meta's ambitions around agentic AI—systems designed to act on behalf of users rather than simply respond to queries—also extend beyond the business sector. The company plans to offer "personal AI agents" and AI-powered smart glasses to consumers.
Furthermore, Meta is actively utilizing its AI technology, including large language models, to accelerate the development of its social applications. Recent examples include new apps for Marketplace sellers, Facebook Groups, and various gaming experiments, with more anticipated releases in the near future.
“I expect it to become a lot easier to ship new apps,” Zuckerberg stated. He anticipates building out more applications and employing recommendation systems to reach relevant user bases more efficiently.
