Editorial illustration for AI Chatbots Lead Hikers Astray, Search and Rescue Veterans Say
AI Chatbots Mislead Hikers, Rescue Teams Warn
AI Chatbots Lead Hikers Astray, Search and Rescue Veterans Say
Three hikers stuck on Mount Shasta called for rescue last summer after a day that went wrong in almost every way it could. They skipped enough food, misjudged their own fitness, and pushed on toward the summit at 7:30 pm instead of turning back hours earlier. By the time they reached the top, descending in daylight was no longer an option. One of them banged up his knee scrambling down in the dark, and they ended up stranded on a rocky stretch of the mountain until help arrived.
Nick Meyers, who runs the US Forest Service's Mount Shasta Wilderness Climbing and Avalanche Center, led the response team that found them. He's used to pulling unprepared hikers off the mountain. What stood out this time was where their bad planning came from: Google's Gemini chatbot, which the group had leaned on instead of doing their own research. Meyers says it's part of a pattern he's seeing more often, from campers citing AI for permission to break posted rules to hikers carrying AI-generated maps that put mountains on the wrong side of the trail entirely.
Turning to an AI chatbot for guidance does come with unique risks. Chatbots can “hallucinate” persuasively presenting misinformation as real, and their use can lead to a phenomenon called “cognitive surrender,” said Chris Flathmann, a human-centered computing professor at Clemson University in South Carolina who has studied AI and search and rescue.
Why this matters
The Mount Shasta case fits a pattern search and rescue teams have seen before: a new information tool, a wave of overconfident users, a spike in rescue calls. Instagram did this a decade ago by selling destinations stripped of context. Chatbots are doing something arguably worse, because they package bad advice in the same confident, authoritative tone they use for good advice.
A chatbot doesn't know a hiker's fitness level, hasn't checked the afternoon thunderstorm forecast, and can't tell the difference between a maintained trail and a scramble that requires crampons. For builders working on consumer AI products, this is a concrete liability problem, not an abstract one: when your interface mimics the certainty of a human expert, users will extend it the same trust, even in domains where the model has no real situational awareness. If your chatbot gets asked for route advice, medical triage, or anything with physical consequences, the honest answer is often "I don't know, check a local source," not a fluent, plausible-sounding paragraph.
The rescue teams quoted here aren't anti-technology. They're describing what happens when confident output meets zero accountability for being wrong.
Related Reading
LLMs & Generative AI ChatGPT's 'Nerdy' tweak rewards goblin metaphors in answers, study finds
LLMs & Generative AI Google to End Free Access to Gemini AI for Non-Subscribers
LLMs & Generative AI AI Engineers Face Rising Costs, Need New Strategies for Efficiency
From Archives ChatGPT's Chat and Work merger signals end of all-you-can-prompt pricing
From Archives Anthropic's Claude Watermark Evaded by Minor Prompt Tweaks
Common Questions Answered
What specific risks do AI chatbots pose to hikers seeking trail guidance?
AI chatbots can "hallucinate" by persuasively presenting misinformation as factual information, creating dangerous situations for outdoor enthusiasts. Additionally, relying on chatbots can lead to "cognitive surrender," where users uncritically accept the chatbot's confident advice without considering their own fitness level, weather conditions, or other critical safety factors that the AI cannot assess.
How did the Mount Shasta hikers' situation demonstrate poor decision-making on the trail?
The three hikers made multiple critical errors including skipping adequate food, overestimating their own fitness level, and continuing toward the summit at 7:30 pm instead of turning back earlier to descend in daylight. These decisions ultimately led to one hiker injuring his knee while scrambling down in the dark, and all three becoming stranded on a rocky section of the mountain until rescue arrived.
Why are chatbots more dangerous than previous information tools like Instagram for outdoor recreation?
While Instagram stripped destinations of context by showing only appealing photos, chatbots present bad advice using the same confident and authoritative tone they use for good advice, making it harder for users to distinguish between reliable and unreliable information. Chatbots also lack critical contextual knowledge such as a hiker's fitness level, current weather forecasts, or terrain-specific hazards that are essential for safe outdoor planning.
What pattern have search and rescue teams observed with new information tools and rescue call spikes?
Search and rescue veterans have identified a recurring pattern: when new information tools emerge, they attract overconfident users who rely on them without proper verification, which subsequently leads to a spike in rescue calls. This pattern has been seen with Instagram a decade ago and is now being repeated with AI chatbots providing hiking guidance to unprepared or inexperienced outdoor enthusiasts.
Further Reading
- ChatGPT's hiking advice left two hikers stranded on a mountain in Poland - Yahoo Tech
- AI directions leading to increasing calls for wilderness rescues - Taipei Times
- Misleading ChatGPT advice leaves hikers stranded - Cybernews
- Think Twice Before Letting AI Decide Your Hiking Route - Yahoo Travel
- Like, follow the river or something: AI chatbots mislead hikers, rescuers warn - GetCoAI
