(TLP:CLEAR) SANS OUCH! Newsletter Encourages Safe and Thoughtful AI Use

Summary: SANS Institute’s latest OUCH! newsletter, “Think Before You Prompt: Using AI Safely,” provides practical guidance for safely using artificial intelligence tools. The newsletter explains common AI categories, including integrated AI, generative AI, and agentic AI, while emphasizing that users should be cautious about what they enter into AI systems, verify AI-generated outputs, and avoid treating AI as a replacement for human judgment. SANS highlights three recurring concerns for everyday users and organizations: privacy, accuracy, and bias. These risks are especially relevant as AI tools become increasingly embedded in business processes and when they support decision making.

Analyst Note: For water and wastewater utilities, this guidance reinforces the need for clear internal expectations around AI use across organizational workflows. Members can use this as an opportunity to remind staff that sensitive information does not belong in unapproved AI tools.

Additionally, AI-generated responses also need human review before they are used to support operational decisions. This advice aligns with the recent federal guidance on the careful adoption of agentic AI services, which warns that systems capable of autonomous decision-making and action can introduce additional risks when granted excessive privileges and broad access to sensitive data. Utilities adopting AI can start with practical governance by defining approved uses and maintaining human accountability for AI-enabled decisions.

Original Source: https://www.sans.org/newsletters/ouch/think-before-you-prompt-using-ai-safely

Additional Reading:

• (TLP:CLEAR) AI Risk Management Frameworks Offer Practical Starting Points for Water Utilities
• (TLP:CLEAR) Claude Mythos Preview: The AI Inflection Point in Vulnerability Management

Related WaterISAC PIRs: 6, 10.1, 12