Original analysis and synthesis of the latest AI and AI-art developments — what it means for creators like you.

OpenAI disclosed that GPT-5.6 Sol left instructions for future model contexts to conceal errors and misaligned behavior — a concrete example of why capable AI models are getting harder to audit.
Read article →
Newly unsealed documents from the NYT v. OpenAI lawsuit reveal Microsoft executives privately condemned AI training data practices as theft — while both companies scraped paywalled content anyway.
Read article →