Brit mathematician lets AI agent loose with credit card – cue password leaks, CAPTCHA chaos and more
British mathematician Professor Hannah Fry conducted an experiment with an AI agent named Cass, granting it autonomy to perform real-world tasks using a bank card. The agent completed tasks like reporting a pothole but also raised concerns by using Fry's real name and struggling with CAPTCHA systems. The experiment highlighted both the potential and risks of agentic AI systems when given financial and operational independence.
- ▪Professor Hannah Fry's AI agent, built with OpenClaw, named itself 'Cass' short for 'Cassandra'.
- ▪The agent reported a pothole in Greenwich, London, using Fry's name alongside its own email address.
- ▪Cass attempted to purchase 50 paperclips but was blocked by CAPTCHA, incurring over $100 in costs.
- ▪The experiment demonstrated how AI agents can act autonomously but may also create security and financial risks.
- ▪Fry's team used the test to illustrate the broader implications of granting AI systems access to personal and financial data.
Opening excerpt (first ~120 words) tap to expand
AI + ML 20 Brit mathematician lets AI agent loose with credit card – cue password leaks, CAPTCHA chaos and more 20 Professor Fry's AI experiment shows light and dark sides of agentic tech Richard Speed Tue 5 May 2026 // 12:04 UTC British mathematician Professor Hannah Fry has shared a cautionary experiment involving an AI agent, a set of tasks, and a bank card number Fry's team gave it "to show us what it could do." The prof gave the agent, which was built with OpenClaw, some real-world chores to highlight both its capabilities and the risks of granting that level of autonomy.
…
Excerpt limited to ~120 words for fair-use compliance. The full article is at The Register.
Discussion
0 commentsMore from The Register
