In June, Meta ventured into the enterprise AI sector with a new AI tool designed to assist businesses in customer service and operational tasks. However, according to CEO Mark Zuckerberg’s comments during the second-quarter earnings call, the company’s aspirations in enterprise AI extend well beyond this initial offering.
Zuckerberg stated, “We recognize a significant opportunity in selling to enterprises, which includes APIs and business agents, as well as potentially offering direct computing services and various solutions we are developing for large clients.” Such initiatives could open doors to new revenue avenues apart from their predominant advertising income and smaller subscription revenues.
Initially, Meta plans to serve its current advertising partners by introducing AI agents compatible with messaging platforms and more. This will enable companies to engage with their customers through an AI-based interaction.
He elaborated, “Similar to our advertising system, we will earn revenue by delivering positive outcomes for these businesses.” Zuckerberg views this strategy as a natural extension of the sales and affiliations Meta has cultivated with millions of advertisers and numerous small businesses using its platforms.
Zuckerberg also highlighted the potential for Meta to reach beyond its existing small business clientele by eventually offering the company’s internal tools to external clients.
“We believe there are larger enterprise clients we can begin to support,” he shared. “We are developing internal productivity tools out of necessity, ensuring they are tailored for our needs. Now that we have these capabilities, we believe there is considerable potential for servicing both small and large businesses.”
This pivot in focus may present challenges, as Zuckerberg acknowledged that selling to enterprises requires different expertise than what Meta is accustomed to.
Regarding the sale of computing resources to enterprise clients, Meta aims to balance immediate revenue needs with long-term strategic goals. The company has indicated multiple times that it has the chance to sell computing power at a substantial markup from its acquisition costs.
Zuckerberg warned investors against the folly of “selling all compute for quick profits.” He described Meta’s strategy as a balanced “portfolio” that integrates both short-term and long-term objectives for its computing infrastructure. “As we advance towards personal superintelligence, we will require hardware that facilitates seamless interaction,” he added.
The call also included discussions about Meta’s ambitious plans for agentic AI—intelligent systems that perform tasks on behalf of individuals or organizations, rather than merely responding to inquiries. These offerings won’t be limited to businesses; consumers can also expect “personal AI agents” and AI smart glasses designed to engage with their surroundings.
Moreover, Meta is leveraging AI technology, particularly large language models, to accelerate the development of its range of social applications. Recent introductions include tools for Marketplace vendors, enhancements for Facebook Groups, and engaging games, with more innovations on the horizon, as teased by Zuckerberg.
“I anticipate that launching new apps will become significantly more straightforward,” Zuckerberg expressed. “Our plans include broadening our range of ideas and utilizing our recommendation systems to reach audiences that will find them appealing.”
