COLIBRIX ONE × BitGN: New Benchmark Reveals AI Reliability Gap

The Fintech Times covered the full findings of ECOM1, the open benchmark from COLIBRIX ONE and BitGN that stress-tested autonomous AI agents against real ecommerce scenarios with over 1,000 engineers. While top architectures neared a 95% success rate, the average run scored just 20.2%, highlighting a reliability gap the article frames as the real barrier to scaled autonomous commerce. COLIBRIX ONE and BitGN are now moving into ECOM2 with fintech-specific scenarios and new partners.