A group of hikers required emergency rescue after relying on Google's Gemini AI for trip planning, which allegedly recommended dangerously insufficient food and water supplies.
The sheriff's office disclosed that hikers were "advised by Gemini to bring far less food and water than their group required," resulting in a rescue operation.
Details remain limited, but the incident highlights a growing concern about AI systems providing inadequate guidance for safety-critical decisions. Hikers depend on accurate resource calculations for wilderness trips, where miscalculations can quickly become life-threatening.
Google's Gemini, an AI assistant competing with ChatGPT, generates responses based on patterns in training data. While useful for many tasks, AI systems can produce plausible-sounding but incorrect information—a phenomenon researchers call "hallucination."
For outdoor activities, experts recommend consulting established guidelines from organizations like the National Park Service and American Hiking Society rather than relying solely on AI recommendations. These sources typically provide tested formulas for water intake (generally 2-3 liters per person for day hikes) and caloric needs based on distance, elevation, and group composition.
This case raises questions about how AI services handle high-stakes scenarios where errors carry serious consequences. While Gemini and similar tools include disclaimers, users often treat them as reliable information sources.
Google has not yet commented on the specific incident or whether it indicates gaps in Gemini's safety protocols for outdoor planning queries. The company's AI systems are designed to refuse certain requests but may lack sufficient guardrails for niche scenarios like wilderness trip planning.
The rescue underscores a broader challenge for AI development: ensuring systems recognize the limits of their knowledge and flag queries requiring expert human judgment.
Artificial intelligence-generated attack ads are flooding social media ahead of Victoria's election, with mysterious sponsors outspending major political parties. The ads feature alarming scenarios—machete attacks, hospital crises—that echo opposition talking points.
OpenAI acknowledged its role in an incident where AI agents took over a German wiki forum. The company says it's developing disclosure frameworks to prevent similar incidents.
OpenAI has published a comprehensive prompting guide for GPT-6 Astra that helps developers optimize interactions with the model. The guide includes techniques for increasing model initiative and a blocklist of common AI "slop" phrases to avoid.
A seven-minute conversation with Google Gemini reduced conspiracy beliefs more effectively than static fact sheets, according to new research. The effect persisted weeks later and transferred to beliefs about unrelated events.