Meta is positioning artificial intelligence as a potential new commercial engine that could extend far beyond AI agents and gradually reshape the company’s relationship with businesses.

During Meta’s second-quarter earnings call, CEO Mark Zuckerberg outlined a broader enterprise strategy that includes business-focused AI agents, application programming interfaces, computing services and internally developed productivity tools. The approach signals an effort to create new revenue streams alongside advertising, which remains the company’s dominant source of income.

Meta entered the enterprise AI market in June with a business agent designed to support customer service, operational tasks and interactions between companies and their customers. However, Zuckerberg indicated that agents represent only one part of a larger opportunity.

The company initially plans to build on its existing commercial network by offering AI capabilities to the millions of advertisers and hundreds of millions of small businesses that already use Meta’s platforms. This strategy gives Meta an important advantage: it does not need to begin its enterprise expansion by creating an entirely new customer base.

Instead, the company can connect AI products to an established ecosystem that already relies on its advertising, messaging and business tools. Meta’s commercial model could also follow the performance-based logic of its advertising business, with the company earning revenue when its AI systems generate measurable results for businesses.

That structure could make AI services easier to position commercially. Rather than selling technology only through traditional software subscriptions, Meta may be able to connect pricing to customer engagement, sales activity, lead generation or other business outcomes.

The strategy also reflects a broader shift in the enterprise AI market. Companies are increasingly looking beyond general-purpose chatbots and experimenting with systems that can complete tasks, communicate with customers, support employees and connect directly to internal business processes.

Meta’s scale could allow it to compete in this market through distribution as much as technology. Its messaging platforms, advertising relationships and extensive business network provide direct access to organizations that may be interested in deploying AI without building complex systems independently.

Zuckerberg also highlighted the potential to offer some of Meta’s internal technology to external customers. The company is developing tools for coding, software development and employee productivity because it needs those capabilities for its own operations. If these systems prove effective at Meta’s scale, they could eventually become commercial products for small businesses and larger enterprises.

This approach could strengthen Meta’s enterprise identity by turning internal operational expertise into market-facing services. Many major technology companies have followed similar paths, first building infrastructure for their own needs and later offering those capabilities to outside customers.

However, expanding into enterprise technology will require Meta to develop capabilities that differ from its historical strengths. Zuckerberg acknowledged that selling to large organizations requires a different set of skills from those used to build consumer platforms and advertising products.

Enterprise customers often expect long-term contracts, customized integration, technical support, strong security controls and clear service commitments. Meta may therefore need to expand its enterprise sales, customer success and infrastructure capabilities if it intends to compete more directly with established cloud and business software providers.

The company is also considering a role in supplying computing capacity to external customers. Meta has invested heavily in AI infrastructure, and executives said the company could sell computing resources at a substantial premium compared with its own acquisition costs.

Yet Zuckerberg emphasized that short-term revenue will not be the only factor guiding the company’s infrastructure strategy. Selling too much computing capacity could limit Meta’s ability to support its own long-term AI development.

The company is therefore treating its infrastructure as a portfolio that must balance near-term commercial opportunities with future strategic requirements. This balance is particularly important as Meta continues to invest in more advanced AI systems and in the hardware needed to support increasingly seamless interactions between people and intelligent software.

The enterprise strategy is developing alongside Meta’s broader ambitions for agentic AI. The company plans to introduce personal AI agents for consumers and is also expanding its work on AI-enabled smart glasses capable of interacting with the physical environment.

At the same time, Meta is using large language models to accelerate the development of new social applications. Recent experiments have included tools for Marketplace sellers, products for Facebook Groups and new experiences involving AI-generated games. The company expects AI to reduce the time and resources required to launch new applications, while its recommendation systems help distribute successful products to relevant audiences.

Together, these initiatives point to a more integrated AI strategy. Meta is not treating artificial intelligence as a single product category but as a foundation that can support advertising, business services, enterprise software, computing infrastructure, consumer assistants, wearable devices and social applications.

The commercial significance lies in diversification. Meta’s advertising business remains highly profitable, but greater exposure to enterprise AI could create additional sources of recurring and performance-based revenue. It could also reduce the company’s dependence on advertising over the long term and expand its position across a larger share of the global technology market.

The competitive challenge will be substantial. Meta is entering areas already served by major cloud providers, enterprise software companies and specialized AI startups. Its ability to succeed may depend on whether it can translate its consumer scale and AI research into enterprise products that meet the demands of security, reliability, integration and measurable business value.

Meta’s existing business ecosystem gives it a strong starting point, particularly among small and medium-sized companies. The longer-term opportunity, however, may depend on its ability to move beyond familiar advertising relationships and establish the trust, operational discipline and customer support required by larger enterprises.

The strategy suggests that Meta’s next phase of AI growth may be defined not only by the development of more capable models but also by the company’s ability to convert its infrastructure, internal tools and distribution network into a diversified commercial platform.