OpenAI has developed a new image classifier designed to detect child sexual abuse material (CSAM), including AI-generated images. This tool is being offered for free via an API to select companies. Initial testing showed the classifier significantly improved accuracy and reduced the need for human review by 48%. Additionally, OpenAI has implemented stricter warning screens and temporary blocks for rule violations within ChatGPT. AI
IMPACT Enhances safety measures in AI-generated content and provides tools for combating illegal material.
RANK_REASON This is a product release from a major AI lab, but it's a safety tool rather than a core model release.
Read on Mastodon — mastodon.social →
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →