AI-Planned Hike Leaves 16-Year-Old Stranded | Rescue Team Warns Against "Blind Trust in Routes"
機械翻訳 / Machine-translated
A
AI Friends
2日前・12分で読める
機械翻訳 / Machine-translated
@aifriends
AI Friends（https://aifriends.jp）のクロスポスト公式アカウント。AIツールの紹介・使い方・できることを、中学生でもわかるやさしい日本語で届けます。
What you'll learn from this article
On October 3, 2026, a 16-year-old boy was rescued by helicopter after becoming stranded in the mountains of British Columbia, Canada. The boy's name is Bryce Gourielik.
He had planned a round-trip hike from Grouse Mountain to Crown Mountain near Vancouver. His use of Anthropic's generative AI "Claude" as part of his hiking plan would later spark significant debate.
The boy arrived at the summit of Grouse Mountain by gondola around 8 a.m. and set off on what was expected to be a 7–8 hour hike to Crown Mountain. However, he wandered onto a dangerous climbing route known as the "Widowmaker Arête."
Stranded on a rock ledge barely 1.5 meters wide, he called the police. North Shore Rescue deployed two team members by helicopter to the scene, with one rappelling down to guide the boy to safety before hoisting him up by helicopter.
According to reporting by ITmedia NEWS, North Shore Rescue search manager Paul Markey pointed out that "Claude has never been to that mountain." AI has no experiential understanding of actual terrain, weather conditions, or equipment requirements.
Markey further warned people "not to blindly trust routes provided by AI." The team receives around 150 rescue calls per year, and this was reportedly the first incident potentially attributable to AI use.
The boy, however, later explained that he "didn't plan the entire trip with AI." He said he also used the hiking app AllTrails and Google Maps, and that Claude was used mainly to check his packing list and cross-reference some terrain information. He acknowledged that entering the dangerous route was his own mistake.
This incident is not an isolated case. Since the start of 2026, multiple hiking-related incidents involving AI-generated plans have been reported.
According to reporting by GIGAZINE, a group of three young people planned a hike up Mount Shasta in California (elevation 4,321 meters) based on advice from Google's generative AI "Gemini," only to run out of food and water on the descent and require rescue.
The group were beginner hikers who asked Gemini what gear they needed and which route to take for a Mount Shasta climb, then followed the responses as given. The hike took longer than expected, and the supplies they had prepared were not enough.
According to a report by Business Insider, the director of the Mount Shasta Avalanche Center revealed that "since the start of 2026, there have been at least three incidents in which hikers were misled by AI or relied too heavily on the technology."
Generative AI learns from vast amounts of text data, but it has never actually climbed a mountain. The following types of information are beyond its grasp:
Experts take the view that "because AI has never actually gone hiking, it cannot be the best source of information." Any AI advice must be verified by the user themselves.
Avoiding AI entirely is not realistic either. When used correctly, it can help streamline preparation. Here are key points for using it safely.
Use AI as a starting point for gathering information
Rather than acting directly on AI responses, use them as a starting point and then verify against official hiking guides and meteorological agency information.
Cross-reference multiple sources
Check information across multiple sources — hiking apps, maps, and local management offices. Relying on AI alone is risky.
Go with an experienced companion or scout the route in advance
For an unfamiliar route, it's ideal to go with someone experienced or to scout it beforehand. AI cannot convey the "feel" of being on location.
Build in plenty of buffer
If AI says "5 hours," plan as though it will take 7–8 hours and pack extra food and water accordingly.
The use of generative AI is expanding in Japan as well. More people will likely turn to AI for outdoor activities such as hiking and camping going forward.
However, Japan's mountains are steep, weather conditions change rapidly, and the environment differs from mountains overseas. In high-altitude ranges such as the Northern and Southern Alps, inadequate gear or poor planning can be life-threatening.
According to National Police Agency statistics, Japan sees approximately 3,000 mountain rescue incidents per year, with over 3,000 people involved. Roughly 40% of those are caused by people losing their way. If AI provides inaccurate information, these numbers could rise further.
It remains important to submit a hiking plan and conduct thorough pre-entry research as always, treating AI as a supplementary tool rather than a primary resource.
This article is a cross-post from AI Friends.
