Generative AI can improve productivity while also creating risks involving bias, privacy, security, and misinformation.

Optimising content for AI discovery is not a purely technical exercise. The systems doing the reading can produce useful summaries, yet the research literature also points to bias, misuse, misinformation, privacy concerns, and security threats. That makes accuracy and accountability part of the process. Content should be easy to quote, but it should also be supported by evidence, kept current, and checked by people.