How OpenAI is defending ChatGPT Atlas from attacks now - and why safety's not guaranteed

An 'automated attacker' mimics the actions of human hackers to test the browser's defenses against prompt injection attacks. But there's a catch.

WebRAT malware spread via fake vulnerability exploits on GitHub

The WebRAT malware is now being distributed through GitHub repositories that claim to host proof-of-concept exploits for ...

Your AI browser can be hijacked by prompt injection, OpenAI just patched Atlas

OpenAI says it has patched ChatGPT Atlas after internal red teaming found new prompt injection attacks that can hijack AI ...

OpenAI Says Prompt Injections a Challenge for AI Browsers, Builds an Attacker to Train ChatGPT Atlas

OpenAI says prompt injections remain a key risk for AI browsers and is using an AI attacker to train ChatGPT Atlas.