Visa 2026 Strategy: AI to Boost Transaction Volume Through Innovation

Summary: Visa’s 2026 plan to use AI spending and capital investment to lift transaction volume and commerce

Based on Visa’s discussion in its 2026 Q2 earnings materials, Visa’s approach is not framed as “AI spend increases volume directly through marketing only.” Instead, management positions AI (and the investments behind it) as improving (1) transaction authorization, (2) fraud/risk capture, (3) merchant and B2B payment automation, and (4) the enablement of new commerce channels (especially agentic commerce and micro-transactions)—all of which are expected to expand activity on Visa’s rails and increase reliance on Visa credentials/tokens. ^123453456

---

1) How AI-driven products are expected to increase transaction activity (especially VAS-linked transactions)

Visa emphasizes that its Value-Added Services (VAS) are largely linked to transactions, cards, and accounts, so improvements in authorization rates, fraud outcomes, dispute resolution, and risk decisions can support more transaction throughput and higher engagement. ²⁴⁷

Key AI-related mechanisms mentioned:

• AI for fraud/risk capture inside VAS: Visa says its Visa Large Transaction Model is being used as a foundational model for AI-powered fraud and risk services, and that early results show up to a 5x increase in fraud value capture. ²
• AI-enabled features integrated across VAS: Visa references integrating AI-enabled features such as Smarter Stand-In Processing and Visa Provisioning Intelligence, and rolling out 6 dispute resolution capabilities—with client adoption fastest among AI embedded services. ²
• VA S growth as the engine tied to commerce: Visa states VAS represents 30% of net revenue and is growing at 25%+ in constant dollars, with durability reinforced “in an AI world.” ⁴

Implication for 2026 transaction volume/commerce: Even though the excerpt does not provide a “2026 AI capex budget” or a numeric forecast specifically tied to AI spend, management’s stated operating logic is that AI improvements in risk, fraud, authorization, and dispute workflows should reduce friction and improve approvals, supporting more transactions and greater commercial/eCommerce engagement through VAS-linked offerings. ²⁴⁷

---

2) “Agentic commerce” as the commerce channel Visa expects AI to unlock—and the role of AI in transaction creation

Visa explicitly connects AI + agentic commerce to increasing its addressable market and creating significantly more transactions, including micro transactions and event-by-event spending.

Management’s articulated expectations include:

• New category of micro transactions where “agents will pay for their own data and resource consumption transaction by transaction and event by event,” creating “an entirely new category of commerce with micro transactions.” ¹
• Accelerated digitization of B2B payments: AI agents are expected to automate payment initiation directly from invoices and contracts, manage approvals autonomously, and drive use of virtual cards and tokenization. ¹
• Agents splitting purchases into multiple optimized transactions (pricing/timing/value). ³
• Visa states it plans to enable CLI commerce at scale (Visa CLI as a platform for developers/agents to pay via Visa credentials), and positions card acceptance as beneficial across ticket sizes including “micro transactions.” ⁶⁵

Implication for 2026 transaction volume/commerce: Visa is essentially betting that AI will (a) increase transaction count by design (agent splitting + micro/event-by-event payments) and (b) reduce friction by automating initiation and approvals—thereby expanding commerce volume that flows through Visa’s network and tokenized credential infrastructure. ¹³⁶

---

3) Capital expenditures / infrastructure spending angle: network scale and “tokenization + trust” as the enabling platform

The excerpts do not directly quantify Visa’s AI-related capex or 2026 capital plan. However, management links the expected growth in agentic commerce to the platform Visa provides—its scale, security, trust, and tokenization—suggesting that capital/investment is meant to maintain and extend network capabilities that agentic commerce will require.

Supporting points from the excerpt:

• Visa frames the differentiator for agentic commerce as trust, not compute/data center capacity. Management says the “limiting factor… is trust,” and emphasizes Visa cards’ ease of use, broad acceptance, privacy, security protections, issuer KYC-backed credentials, and liquidity management advantages versus alternative approaches. ⁸
• Visa provides network scale and credential reach as the foundation: more than 175 million seller locations, 5 billion credentials, nearly 14,500 financial institution clients, and over 300 billion transactions annually (about 900 million per day). ¹
• Tokenization is positioned as foundational in the agentic world, including preference for cards and the role of Visa tokens. ⁵

Implication for 2026 transaction volume/commerce: While capex numbers aren’t given, Visa’s discussion implies that investments (whether capital or ongoing platform spend) are aimed at sustaining and extending the network/security/tokenization layer that agentic commerce will rely on for trusted payments—thereby supporting higher transaction throughput and adoption of Visa credentials in new commerce flows. ¹⁵⁸

---

4) Evidence that Visa’s strategy is already affecting commerce activity (current transactions and adoption signals that may scale into 2026)

Visa provides some current-quarter results and adoption proof points that reinforce its thesis that AI and solution improvements drive commerce engagement:

• In Q2, Visa reports payments volume up 9% (constant dollars) to $3.7 trillion and processed transactions up 9% to $66 billion (note: the excerpt states processed transactions grew to $66 billion; typically this refers to transaction count—the excerpt’s wording should be read as presented). ⁴
• Visa states VAS revenue growth and adoption: VAS growth at 27% year-over-year in constant dollars to $3.3 billion, with growth tied to process transactions, credential mix/number, and consulting/marketing engagements plus pricing. ⁷²
• For specific commerce/enablement layers, Visa points to early real-world “agentic commerce” transactions via tokens and references early growth in agentic shopping. ⁵
• Management also notes partnerships enabling digitization and automation in commercial flows (e.g., expanding travel virtual cards and automation/control/reconciliation through “Visa Commercial Choice for Travel”). ³

Implication for 2026: These excerpts don’t provide a direct 2026 bridge from AI capex to volume, but they do show management expects the same AI-enhanced authorization/risk and the same digitization/agentic enablement themes to scale—especially where VAS adoption is fast among AI-embedded services. ²⁴³

---

5) Direct answer to the “how” question: Visa’s planning levers

Putting the excerpted logic together, Visa’s plan to use AI spending and capital investment to impact transaction volume and commerce in 2026 centers on these levers:

1. Increase authorization and reduce failure/fraud friction through AI risk/fraud models integrated into VAS, supported by claims like the Visa Large Transaction Model enabling up to a 5x increase in fraud value capture. ²
2. Grow VAS because it is transaction-linked, so improvements propagate to more transactions/carding/accounts, with VAS already at 30% of net revenue and 25%+ growth in constant dollars. ⁴²
3. Unlock agentic commerce transaction growth by enabling trusted payments for micro/event-by-event transactions and B2B automation (agent-driven initiation, autonomous approvals), including increased usage of virtual cards and tokenization. ¹³⁶
4. Enable scalable commerce interfaces (Visa CLI and standards) so agents/developers can pay through Visa credentials at scale—positioned as advantageous for both regular and micro transactions. ⁶⁵
5. Maintain the trust/security/tokenization platform as the prerequisite for agents to transact, explicitly framed as the key limiting factor for agentic commerce. ⁸⁵¹

---

What the excerpts do not specify

• They do not provide a numeric 2026 AI capex amount, a quantified “AI spend → transaction volume” causal model, or a specific 2026 transaction-volume target attributable to AI investment. ^12345864

---

Conclusion

In Visa’s own framing, AI spending and related platform investment are intended to expand commerce by increasing the transaction opportunities created by AI agents (more—often smaller—transactions) and by reducing friction through AI-enabled authorization/fraud/risk and automated payment workflows. ¹³² Visa expects agentic commerce to scale through Visa’s trust, security, acceptance, and tokenization infrastructure, with VAS serving as a transaction-linked monetization layer that benefits from AI-driven risk/fraud value capture. ⁸⁵⁴²