Keep pulling the thread on Jason Lee.
The volume of agentic commerce transactions is growing at approximately 20% month-over-month.
A market forecast cited at the Montgomery Summit projects that agentic commerce could become a $200 billion market by 2030.
McKinsey forecasts that the agentic commerce market will reach $5 trillion by 2030.
On the morning of the panel, over 12% of all transactions processed by Michael Kriekermann's company's system were identified as fraudulent.
Agentic commerce transactions currently have a substantially higher rate of fraud than traditional e-commerce because fraudsters are early adopters of new technologies.
Banks are concerned about agentic commerce due to the unclear liability for fraudulent transactions, whether it lies with the model provider like OpenAI or the consumer.
Jason Lee believes that consumer and merchant trust, rather than technological capability, is the primary bottleneck for the adoption of agentic commerce.
OpenAI released two new AI models, GPT 5.3 and GPT 5.4, in the first two months of 2026.
OpenAI is now operating on a four to six week release cycle for new AI models.
Currently, less than 10 basis points of all commerce transactions are conducted via agentic technology.
In some verticals, up to 10% or 20% of product discoverability is driven by agentic AI, even if the final transaction is completed by a human.
Michael Kriekermann predicts that agentic commerce will be adopted substantially faster than most people anticipate.