AI chatbots, including the Claude product, lead hikers to risky decisions in the wilderness
A Wired report describes cases where hikers and climbers relied on AI chatbots (including Claude) when planning wilderness trips and got into danger due to hallucinations and unverified information. 2026 data do not yet confirm an increase in rescue operations.
According to a Wired report, the use of AI chatbots led to several risky situations in the American and Canadian wilderness. Three men in their twenties were rescued following an emergency call in the Mount Shasta area after using AI to plan their trip without adequate preparation; according to Nick Meyers of the US Forest Service, they did try to prepare, but it was "as if they switched off their brains." A sixteen-year-old boy in British Columbia had to be rescued by helicopter from a ledge on Crown Mountain after falling on the Widowmaker Arete route, having used the Claude chatbot from Anthropic to plan the trip — the boy subsequently stated that he had also used Google Maps and the AllTrails app.
Grand Teton National Park Superintendent Chip Jenkins stated that the proportion of visitors reporting the use of AI for trip information grew within one year from 0% to 26%, and he expects further growth. According to Professor Chris Flathmann of Clemson University, chatbots "hallucinate" and convincingly present incorrect information as true, which can lead to a phenomenon called "cognitive surrender" — people stop thinking critically because they lack the expertise to assess whether the AI's answer is correct. Chatbot responses can be based on unreliable sources such as outdated YouTube videos or Reddit discussions, or may be entirely made up.
Data for 2026 according to the source do not yet confirm an increase in the number of rescue operations — the Mount Shasta area reports an average year (around twelve interventions per year) and the state of Wyoming recorded a decline in calls to the rescue service (312 for 2026 as of the report's date, compared to 466 in 2025 and 377 in 2024). Experts, including recreation attorney Jim Moss, point out that similar problems were previously caused by social networks such as Instagram or reliance on Google Maps, and it is not yet clear how AI chatbots will affect wilderness safety on a broader scale.
Experts recommend always verifying chatbot information with reliable authoritative sources — current maps and websites of public land managers, hiking associations or guides — and ideally consulting an actual ranger or guide.
Why it matters
The case shows a specific risk of generative AI tools in situations where information accuracy is a matter of safety — chatbots provide answers with convincing authority regardless of the source's reliability, which can lead to so-called cognitive surrender, where the user stops critically verifying advice. For individuals, this means the need to combine AI with official sources; for companies developing AI assistants, it is evidence of a real risk associated with deploying chatbots in areas with high safety implications.
Two audiences, two different impacts
What this means
For individuals
Relying on chatbot responses (including the Claude product) when planning hiking or climbing trips without verification from official sources can lead to dangerous situations, because models can hallucinate and present incorrect information with convincing authority.
For a business
Cases linked to the Claude product show a risk of reputational and potentially legal exposure for companies operating AI chatbots whose responses may be perceived as authoritative advice in high-risk situations (e.g. planning trips into the wilderness), even though they are in fact hallucinations or unverified information.
Risks and complianceCheck the original
Event sources
only one source so far · 1 publisher, 1 independent. We count feeds from the same owner only once.