Morph has published a report projecting that AI agent-influenced commerce will exceed USD 500 billion in global gross merchandise value by 2028.
The report, titled The Agentic Economy, examines the emergence of agentic commerce, a category of economic activity in which AI systems can independently discover products, compare prices, negotiate terms, and execute payments on behalf of users. The publication arrives as major payments and technology companies move to build foundational infrastructure for machine-driven transactions.
Spending signals and industry investment
According to the report, AI-influenced spending reached USD 67 billion globally during the 2025 holiday season. Adobe recorded a 693% year-on-year increase in generative AI-driven retail traffic during the same period. McKinsey estimates the long-term market opportunity for agentic commerce could exceed USD 1 trillion in US retail alone, though Morph does not specify a timeline for that figure.
Several companies are already positioning for this shift. Visa has confirmed hundreds of secure agent-initiated transactions across its ecosystem. Google expanded its Agent Payments Protocol (AP2) with more than 60 launch partners and added native stablecoin settlement support. Shopify, Stripe, Mastercard, Coinbase, and OpenAI are also developing infrastructure connected to machine-driven commerce.
Predictions and stablecoin infrastructure
Morph's report includes several forward-looking projections. By 2028, the firm expects agent-influenced commerce to exceed USD 500 billion in global gross merchandise value, AI agents to overtake humans in commercial stablecoin payments, and one in ten US households to regularly delegate purchasing decisions to AI agents.
In addition, the report argues that stablecoins may become central to this environment because machine-driven commerce requires payment infrastructure that is programmable, continuously available, low-cost, and interoperable across borders. Traditional payment rails, designed for human-initiated transactions, may be poorly suited to the speed and volume at which autonomous agents operate.
On the question of consumer trust, the report does not treat it as a prerequisite for adoption. It suggests that consumers may simply become accustomed to delegating minor decisions incrementally, until broader delegation becomes a default behaviour rather than a considered choice. This gradual normalisation, rather than a deliberate shift in attitudes, is presented as the likely adoption pathway.
Developer tooling
In order to support developers building AI-powered applications, Morph has open-sourced a toolkit called Morph Skills. The toolkit is designed to allow AI systems to execute blockchain actions (including wallet operations, token swaps, and stablecoin payments) through natural language instructions, lowering the integration barrier for agent-based payment functionality.
The report positions agentic commerce not as a distant prospect but as an infrastructure question already being answered by the industry's largest participants.
