AI Agents Fail 97% of Real Freelance Tasks, Pouring Cold Water on Autonomy Hype
A viral research finding shows that top AI models failed 97% of actual freelance projects sourced from real marketplaces — a stark contrast to the agent hype cycle playing out across every AI company's marketing.
Researchers tested leading AI models on real freelance projects — the kind of work actual humans get paid to do on platforms like Upwork and Fiverr — and the failure rate was 97%. The finding, shared by @MrEwanMorrison and quickly going viral, is the most concrete data point yet on the gap between agent demos and agent reality. The study reportedly tested across multiple top-tier models, meaning this isn't a single vendor's weakness — it's a category-level limitation.
Unlock the full briefing
Get every story in today's briefing, the full archive, and the daily AI intelligence brief.
All stories today
Full archive
Daily brief
Cancel anytime. Payments powered by Stripe.