AI safety and policy discussion tends to split into technical alignment work and regulatory process, and both belong here alongside the practical ethics questions that affect ordinary users.
Technical threads cover alignment and post-training methods, evaluation and red teaming, interpretability, and the disagreements about which risks warrant which response. Policy threads cover the EU AI Act and comparable frameworks elsewhere, export controls and their effect on model availability, liability, and copyright and training data litigation.
The practical layer matters most to members. Disclosure norms, appropriate use in professional settings, and what to do when a tool's terms change under you are recurring topics with no settled answers.
Members hold substantially different views here and are asked to argue positions rather than characterise opponents. Specific claims, cited sources and stated uncertainty make this category worth reading. See AI & Society for the cultural questions and Privacy & Data for data handling.
Popular Discussions
The most-engaged ai safety, policy & ethics discussions in the WhatAI community.
Hi All, On the important questions we should all be asking about AI, I just watched this powerful (and quite sobering) interview that dives deep into **AI safety and ethics**. **Video:**\ **The AI ...
By whitney
May 24, 2026
AI Safety, Policy & Ethics
✓ Reviewed for community standards
❤ 1
👁 6
💬 3
🚩 Report
I don’t see this discussed enough outside technical circles. Everyone talks about what AI can do, but not nearly enough about what it costs to run at scale. Massive training runs, constant inference,...
By jackwright
Apr 10, 2026
AI Safety, Policy & Ethics
✓ Reviewed for community standards
❤ 0
👁 2
💬 0
🚩 Report
I’ve been going back and forth on this one.🤯 Sometimes I see AI generate writing, music, images, or ideas that genuinely feel creative. Other times it feels more like an ultra-fast pattern machine t...
By FahToDeep
Apr 14, 2026
AI Safety, Policy & Ethics
✓ Reviewed for community standards
❤ 0
👁 1
💬 0
🚩 Report