Is there one official number for AI agent payments market size in 2026?

No. Most public sources provide adjacent indicators and forecasts, so the best approach is triangulation across enterprise AI spend, ecommerce behavior, and payment-network rollout data.

What matters most for 2026 sizing assumptions?

Adoption rates, share of transactions that become agent-assisted, and payment-rail readiness. Those three variables drive most of the variance in market-size outcomes.

AI Agent Payments Market Size (2026): Agentic Commerce Growth Analysis

If you search "AI agent payments market size 2026," you quickly hit a problem: there is no single canonical number yet.

That does not mean the market is unknowable. It means you need to triangulate from public signals that are already strong enough to support planning.

This post uses that triangulation method: enterprise AI spend, ecommerce behavior, network rollout data, and long-range scenario research.

Step 1: Set the spend context

Any market-size estimate for agentic payments sits inside broader AI spending.

Gartner estimated global generative AI spending at $644 billion in 2025, up 76.4% year over year (Gartner).

IDC projects global AI spending will exceed $632 billion by 2028 with a 29.0% CAGR (IDC).

Implication: budgets for AI systems are not a blocker. The gating factor for AI agent payments is implementation trust: identity, authorization, verification, and controllable spend.

Step 2: Confirm user-side demand is real

Consumer-side willingness is not universal, but it is already meaningful.

EY reports 48% comfort with AI-driven product recommendations and 31% comfort with autonomous AI purchases (EY).

This matters for 2026 because market size is not just a technical function of what platforms can do. It is also a trust function of what buyers allow.

Step 3: Confirm commercial behavior is already happening

Salesforce reported $229 billion in 2025 holiday online sales influenced by AI, with AI accounting for 19% of holiday orders in its tracked dataset (Salesforce).

This is the clearest public signal that AI-assisted buying has moved beyond experimentation. When AI influences one-fifth of order flow in a peak season, payment infrastructure stops being hypothetical.

Step 4: Check network-level rollout velocity

Payment networks are deploying dedicated agentic programs.

Visa says Intelligent Commerce now includes 100+ partners, with 30+ actively building in sandbox and 20+ direct integrations (Visa).

Mastercard says one in four online shopping interactions could involve AI agents by 2026, based on its commissioned research published alongside Agent Pay (Mastercard).

These are not final market totals, but they are strong leading indicators: payment-rail owners are explicitly productizing for agent-initiated commerce.

Step 5: Bound the long-run upside

Morgan Stanley's scenario work puts upper-bound context around where this can go: up to $4.4 trillion U.S. AI-assisted ecommerce by 2030, and potentially 52% of U.S. ecommerce by 2035 in aggressive adoption cases (Morgan Stanley).

You should treat this as a scenario envelope, not a base case. But it is useful for strategy because it highlights the possible scale gap between teams that build payment controls early versus late.

A practical 2026 sizing framework

Because public sources are fragmented, the most defensible 2026 model is scenario-based.

Scenario inputs to model

▸Adoption input: % of active users or enterprises with live agent purchasing flows.
▸Commerce input: % of transaction volume influenced or executed by AI agents.
▸Rail input: % of merchants and payment stacks that can reliably accept agentic transactions.
▸Control input: loss rate from failed verification, policy drift, fraud, and disputes.

Directional reading from current public data

▸Adoption is accelerating (McKinsey + Capgemini signals).
▸Commerce influence is already material (Salesforce signal).
▸Networks are building production pathways (Visa + Mastercard signals).
▸Upside is very large if trust infrastructure holds (Morgan Stanley envelope).

That combination supports a "high-growth but uneven" 2026 market thesis.

Where most forecasts go wrong

Many market estimates over-focus on model capability and underweight payment operations. Three operational constraints dominate real growth curves:

▸Verification quality: Can you prove who acted, what was authorized, and why the charge happened?
▸Spend isolation: Can you cap blast radius by agent/workflow?
▸Dispute readiness: Can finance teams explain and challenge transactions quickly?

If these fail, growth stalls regardless of model quality.

What to track quarterly in 2026

If you are operating in this market, track these indicators each quarter:

▸AI-influenced order share.
▸Autonomous completion rate (agent completes purchase without human handoff).
▸Approval-to-decline ratio by policy tier.
▸Dispute rate for agent-initiated transactions.
▸Margin by billing unit (token, call, workflow, completed purchase).

These operational metrics are usually better predictors of real market capture than top-down TAM narratives.

2026 conclusion

So how much will agentic commerce grow?

The most defensible answer today is:

▸Large and fast at the category level (confirmed by enterprise spend and commerce signals).
▸Highly variable by operator (determined by payment controls and verification quality).
▸Potentially massive long-term (supported by upper-bound scenario research).

In other words, 2026 is the transition year where "agentic commerce" moves from trend language into measurable payment infrastructure.

Looking for agent spending controls? Start with MCP + skills, then choose a plan that fits your workload.

How Much Will Agentic Commerce Grow? AI Agent Payments Market Size in 2026