Three hikers were rescued from Mount Shasta in California after relying on Google’s AI chatbot Gemini to plan their expedition, according to a report from the Siskiyou County sheriff’s office. The young men began their hike at 3 a.m. and continued despite guidance advising hikers to turn back if they hadn’t reached the summit by noon. Instead, they made it to the top at 7 p.m., which led to a dangerous nighttime descent and eventually an overnight stay in Mud Creek Canyon.

The sheriff’s report highlighted that Gemini advised the hikers to bring significantly less food and water than necessary for their group, which was a critical factor given that their planned 8-hour climb extended into a multi-day ordeal. This lack of sufficient supplies compounded the risk they faced after they failed to descend before dark. Despite the AI’s involvement in their preparation, it remains unclear how much blame the chatbot carries for the poor decisions made during the trip.

Local authorities emphasized the importance of consulting with the US Forest Service Mount Shasta ranger station before embarking on such hikes. They warned against depending solely on AI tools for trip planning, urging hikers to seek accurate, up-to-date information directly from park officials to ensure their safety on the mountain. The rescue operation was conducted by Forest Service rangers and volunteers who found the trio the following morning.

This incident serves as a cautionary tale about the current limits of AI assistance in real-world, high-risk activities like backcountry hiking. While AI tools like Google Gemini offer helpful information, human judgment and local expertise remain essential components of safe outdoor adventure planning. The sheriff’s office’s statement reflects growing concerns about over-reliance on emerging AI technologies without adequate backup from traditional, reliable sources.