Visa announced 2,600 layoffs across its technology and product organizations on the same day as its fiscal Q3 earnings report, framing the cuts as a consequence of AI automating payments network engineering. CEO Ryan McInerney’s internal memo, as reported by Tech Times, describes AI reshaping how Visa builds and operates its infrastructure. The reduction amounts to roughly 7% of Visa’s workforce and represents the company’s largest restructuring to date. The AI narrative is tidy and defensible with investors, but it papers over a more uncomfortable structural pressure: stablecoin payment rails are steadily eroding the interchange and processing fees that constitute Visa’s core revenue engine, and the company is cutting costs ahead of that margin compression rather than behind it.
The AI Story Is Real but Incomplete
McInerney’s attribution of the layoffs to AI automation is not dishonest. Payments network engineering genuinely involves repetitive, automatable work — fraud detection model tuning, transaction routing optimization, compliance monitoring, and reconciliation logic are all areas where large language models and ML pipelines can replace headcount. The problem is that AI automation alone does not explain why Visa is cutting now, at this scale, and in technology specifically. Visa’s fiscal Q2 showed 17% revenue growth and $3.7 trillion in payments volume, per TradingNEWS consensus data. A company growing at that clip does not typically shed 7% of its workforce unless it expects a structural change to its cost-of-revenue trajectory. Stablecoin payment volume — growing on Base, Solana, and other chains — settles without Visa’s network and without the 1 to 3% interchange fees that fund Visa’s operating model.
Stablecoin Rails Are the Actual Disruption
The Tech Times reporting on the layoffs includes a secondary angle that gets closer to the truth: stablecoins are driving the pivot. Payment volume in USDC, USDT, and PYUSD has been scaling across Base and Solana, with Coinbase reporting over 100 million x402-protocol transactions for AI agents alone. None of that volume touches Visa’s network. When an AI agent pays for an API call using USDC through x402, the settlement happens on Base in seconds for fractions of a cent. Visa earns nothing. As agent-to-agent commerce scales — Tiger Research projects a 7x revenue leap from agent wallet infrastructure — the share of digital payments bypassing card networks grows with it. Visa is not facing a sudden collapse in revenue. It is facing a gradual but compounding loss of the high-margin transaction fees that justify its current cost structure.
What the Restructuring Signals
The decision to cut in technology and product, rather than in sales or operations, tells a specific story. Visa is reducing the engineering capacity needed to maintain and extend its legacy rails. That is a bet that the existing network infrastructure requires less incremental investment because the growth in payment volume will increasingly happen on infrastructure Visa does not build. The company is also presumably reallocating some capital toward stablecoin-adjacent strategies — partnerships, issuer relationships, and potentially its own tokenized settlement layer — though no such announcement accompanied the layoff news. The risk is straightforward: if stablecoin adoption accelerates faster than Visa’s new initiatives can capture value, the cost savings from these layoffs will not offset the revenue erosion.
What to Watch
The fiscal Q3 earnings release should reveal whether Visa acknowledges stablecoin-driven disintermediation in its forward guidance or maintains the AI-efficiency framing. Watch for any mention of stablecoin partnerships, tokenized settlement pilots, or changes to interchange pricing. Additionally, the trajectory of agent-driven payment volume on Base via x402 and on competing protocols like Skyfire and Payman will indicate how quickly machine-to-machine commerce is pulling settlement volume off card rails. Visa’s network effects remain formidable — merchant acceptance, consumer habit, and the credit infrastructure that stablecoins cannot yet replicate — but the cost structure is being repriced in real time.
Sources
- https://www.techtimes.com/articles/321858/20260728/visa-cuts-2600-tech-jobs-ai-automates-payments-network-engineering.htm
- https://www.techtimes.com/articles/321833/20260728/visa-cuts-2600-tech-jobs-ai-gets-credit-stablecoins-drive-pivot.htm
- https://www.tradingnews.com/news/visa-362-usd-sits-3-usd-from-its-52-week-high
- https://cryptobriefing.com/stripe-economist-ai-productivity-growth/