Source time Edition date
IndustryWired AI
Chatbots Are Leading Hikers Astray, With Dangerous Consequences
聊天机器人正把徒步者引向歧途，后果危险
Wired warns that hikers who outsource route planning to chatbots risk surrendering the critical thinking that backcountry safety depends on, a tradeoff with potentially deadly consequences.
Judgment
Park superintendents and search-and-rescue managers should decide now whether to treat AI route advice as a trailhead safety issue—adding verification warnings where permits are issued—because the rescue cost falls on public teams while the convenience accrues to chatbot users.
Basis
Nick Meyers of the Mount Shasta wilderness center said the three hikers who relied on Gemini had "turned their brains off": they overestimated fitness, missed a turnaround, summited at 7:30 pm, then descended in darkness and needed rescue. He also described a camper citing AI as permission to camp by a "No Camping" sign and an AI map that put features on the wrong side of the mountain. Clemson's Chris Flathmann names the mechanism "cognitive surrender": an AI that poses as an expert is hard to challenge when the user lacks expertise, and hallucinated directions can be scraped from Reddit or old YouTube videos. Grand Teton's Chip Jenkins said AI travel-info use went from zero to 26 percent. Yet the source's rescue numbers are mixed: Wyoming had 312 calls so far in 2026, versus 466 in 2025 and 377 in 2024, and Meyers's office averages a dozen rescues a year.
- 2024377
- 2025466
- 2026 so far312
Unclear The source does not say how many Wyoming or Mount Shasta rescues involved AI route planning, or whether AI use has changed the accident rate.
Next Next, ask Grand Teton and Mount Shasta officials for 2026 incident reports coded by whether AI was used for route planning.