Visa benchmark finds AI shopping agents fail 42.8% of runs
See how this story connectsVisa launched the Visa Agentic Commerce Benchmark, which put 22 AI models from 12 labs through 15 shopping tasks, and said the models failed 42.8% of test runs in the first round. The best model passed 82.2% of attempts and the worst 28.9%. In 54% of the failed runs the AI still made a purchase, either of a product that did not meet the customer’s needs or before it had found every requirement.
Why it matters
In 54% of failed runs the AI still bought something, and Visa plans to extend the tests to checkout, payment and ordering.
- When
- 5 Nov 2026
- ID
WTA-N-2026-0742- Where
- International
- Who
- Visa; Rubail Birwadker
- Topics
- AI agents, Financial risk
Summary written with AI by wheresthe.ai, from The Next Web. Checked against the source on 5 Nov 2026.
Source copy: saved on 5 Nov 2026 to check this story and not republished, because the text belongs to its publisher. Also held by the Internet Archive.
Cite this story
WTA-N-2026-0742: “Visa benchmark finds AI shopping agents fail 42.8% of runs.” wheresthe.ai, 5 Nov 2026. https://wheresthe.ai/d/intl-visa-benchmark-finds-ai-shopping-agents-fail-42-8-of-ru-2026-11/ (source: The Next Web).
Spotted an error? Email [email protected].