Three hikers were rescued from California’s Mount Shasta this week after using Google Gemini to plan their trip, TechCrunch reports, citing the Chicago Tribune and a Siskiyou County sheriff’s office report. According to TechCrunch’s account of the sheriff’s report, the three young men started hiking at 3 a.m. Hikers are advised to turn around if they have not reached the summit by noon, but the group did not reach the top until 7 p.m. The timing left the hikers trying to descend in the dark. TechCrunch reports that they called the sheriff’s office for directions, then spent the night in Mud Creek Canyon before Forest Service rangers and volunteers rescued them the next morning. The sheriff’s office did not frame Gemini as the only cause of the incident. But it said the hikers had been advised by Gemini to carry far less food and water than their group needed, especially once what had been planned as an eight-hour ascent turned into a multiday ordeal. The office’s takeaway was practical: hikers should call the local U.S. Forest Service Mount Shasta ranger station before a trip to get current information, and should not rely only on AI systems for route or safety planning. For AI providers and users, the case is a reminder that general-purpose chatbots can be used in settings where bad advice can become a physical safety issue. The reported facts do not establish that Gemini caused the rescue, but they do show how AI-generated planning can become part of the chain of decisions in high-risk activities. Who benefits: Local ranger stations and expert trip-planning resources benefit when official guidance is reinforced as the authoritative source for current conditions. AI teams that design stronger safety handoffs for outdoor or travel advice also have a clear use case to address. Who's exposed: Users who treat general-purpose chatbots as sufficient for wilderness planning are exposed to bad assumptions about route time, food, water, and daylight. AI providers are exposed when consumer use cases create safety expectations their products may not be designed to satisfy.