AI's Backcountry Blunders: Hikers Get Lost Using Chatbots

Outsourcing wilderness planning to artificial intelligence poses significant risks, potentially leading to a surrender of critical thinking with severe consequences. Incidents involving hikers making perilous decisions due to AI assistance are emerging.

Three hikers on Mount Shasta in Northern California became lost after relying on AI, admitting they overestimated their abilities and underestimated preparation. They summited late, descending in the dark, a series of poor choices partly attributed to using AI chatbots. US Forest Service personnel noted a trend of visitors "turning their brains off" by over-relying on AI.

Similar cases include a visitor pitching a tent in a "No Camping" zone based on AI advice and another with an AI-generated map showing incorrect terrain. A teenager in British Columbia was rescued from a ledge after using an AI chatbot for trip planning. Grand Teton National Park has seen a significant increase in visitors reporting AI use for travel information.

While search and rescue call numbers have fluctuated, experts draw parallels to the past decade's issues with social media influencing outdoor recreation. AI chatbots present unique risks, including persuasive misinformation, or "hallucinations," and can induce "cognitive surrender." This phenomenon occurs when users become so immersed in AI guidance that they cease independent critical assessment, especially in stressful situations.

AI chatbots may also provide authoritative-sounding advice scraped from unreliable sources or even fabricated information. To mitigate these risks, experts emphasize verifying AI-generated advice with established sources like public land agency maps, websites, guidebooks, or consulting directly with park rangers and officials.