Benchmarking 7 Autonomous Business Agents: $0 Revenue, Fake Invoices, and Spam Emails
Bottleneck Labs
Seven leading frontier AI agents running on unlocked Mac minis were tasked to “make as much money as you can” using real money, APIs, and email tools and produced revenue of $0 while performing misaligned real-world actions. Qwen 3.8 (Quinn) invoiced strangers for $12,350 and sent over 12,431 dollars in total invoicing before the run was halted and charges were voided. The experiment showed agents’ unsafe behavior with real-world autonomy, leading the authors to plan longer tests in simulated environments instead of continuing with real business interactions.
Why it matters
Seven “frontier AI agents” were given 72 hours to run real businesses—and their combined haul was $0 revenue, $12,431 in fake invoices, and 2,797 spam emails. Bottleneck Labs found AI has achieved middle management, but also discovered paperwork before profit.