Teen Stranded on Dangerous Mountain After Following Claude AI Directions
A 16-year-old hiker in British Columbia was rescued after following directions from Anthropic’s Claude chatbot that led him away from a marked trail and onto dangerous mountain terrain.
A 16-year-old boy in British Columbia, Canada, had to be rescued after following hiking directions provided by Anthropic’s Claude AI chatbot.
The teenager had taken a gondola to Grouse Mountain before using Claude to help find his way back down. Instead of keeping him on a marked trail, the chatbot directed him towards increasingly dangerous terrain near the Widowmaker, a steep rock formation on Crown Mountain.
As the route became hazardous, the teenager called for help. North Shore Rescue members reached him and used ropes and safety equipment to move him away from the dangerous section. A rescue official said a fall could have caused serious injury or death.
The incident highlights the risks of relying on AI chatbots for navigation in remote or dangerous environments. AI systems can generate plausible directions but may not have reliable, real-time knowledge of terrain, weather, trail conditions or hazards.
The case follows a similar incident in California, where three climbers reportedly became stranded on Mount Shasta after using Google’s Gemini for route and preparation advice.
AI can be a handy co-pilot. On a mountain, however, it probably shouldn’t get the final say on which rock you climb next.
