
OpenAI integrates Codex into ChatGPT Work, raises security concerns
OpenAI has added Codex to ChatGPT Work, letting the assistant help with reading documents, running code, and editing files inside a workspace. Early reviews suggest companies see faster document and code work, but also worry about security, especially if teams do not control access and monitor use. Research shows that up to a third of prompts may include sensitive data, and there are risks like "shadow AI" use, weak logging, and possible bad advice from the assistant. Experts suggest strict security steps may reduce, but not fully remove, these risks. Whether ChatGPT Work brings lasting benefits appears to depend on strong rules, security controls, and regular checks.













