Meta pivots to agentic commerce in 2026 AI roadmap

Meta users will start to see new AI models and products from the company in a matter of months, with a particular focus on agentic commerce tools. Mark Zuckerberg, CEO of Meta, announced on an investor call on Wednesday that the company plans to ship these new models and products over the coming months. These agentic commerce tools—defined as AI-driven shopping assistants that perform tasks for users—will allow people to find products from the businesses in Meta’s catalog. Zuckerberg stated that the company’s AI is starting to see the promise of understanding personal context, including user history, interests, content, and relationships, to deliver a personal experience.

To support this rollout and its broader AI infrastructure, Meta is significantly increasing its infrastructure spending. The company disclosed its financial projections during its recent quarterly earnings report:

• 2026 Capital Expenditures: Meta anticipates spending between $115 billion and $135 billion over the course of 2026. Capital expenditures represent the funds a company uses to acquire, upgrade, and maintain physical assets.
• 2025 Capital Expenditures: The projected 2026 spending marks a substantial increase from the $72 billion spent in 2025.
• Long-term Infrastructure Projection: The 2026 capital expenditure guidance is part of a larger infrastructure push, with Meta’s total infrastructure spending through 2028 reportedly projected at $600 billion. The company attributed the jump in 2026 spending to increased investment to support its core business and its Meta Superintelligence Labs efforts.

This strategy places Meta in competition with Google and OpenAI, both of which have already built platforms for agent-enabled transactions alongside partners like Stripe and Uber. To bolster its technical capabilities, Meta acquired Manus, a general-purpose agent developer, in December. While other AI developers have established technical infrastructure, Meta is betting that its access to personal user data will differentiate its tools. As Mark Zuckerberg, CEO of Meta, noted on the investor call, “We’re starting to see the promise of AI that understands our personal context, including our history, our interests, our content and our relationships.”

Why it matters

Meta is betting that its unique access to personal user data will allow it to build superior agentic shopping tools, differentiating its AI strategy from competitors like Google and OpenAI.