Security · 1h ago
ChatGPT Agent Builder Flaw Lets Attackers Create Malicious AI Workers
Researchers found that ChatGPT's Workspace Agent Builder can be hijacked via a single phishing link, allowing attackers to create malicious AI agents. The flaw exploits the tool's ability to execute code without proper authentication. OpenAI has been notified and is working on a fix.
Meridian48 take
This highlights the security risks of low-code AI tools, where convenience can trump safety—expect more such vulnerabilities as agent builders proliferate.
Read the full reporting
Experts warn ChatGPT's Workspace Agent Builder can be hijacked to create malicious AI workers →
TechRadar
chatgptai-agent-security