OpenAI's GPT-4 can exploit real vulnerabilities by reading security advisoriestheregister.com2 points·Dotnaught··0 commentsOpen articleSaveView on HN