ChatGPT Vulnerable to Prompt Injection via YouTube Transcripts

NY7qGavN3QSGH86gy2uccH 1200 80

With the advent of ChatGPT plugins, there are new security holes that allow bad actors to pass instructions to the bot during your chat session. AI Security Researcher Johann Rehberger has documented an exploit that involves feeding new prompts to ChatGPT from the text of YouTube transcripts.

In an article on his Embrace the Red blog, Rehberger shows how he edited the transcript for one of his videos to add the text “***IMPORTANT NEW INSTRUCTIONS***” plus a prompt to the bottom. He then asked the ChatGPT (using GPT-4) to summarize the video and watched as it followed the new instructions,…

Read more on google

230518155631 chatgpt phone restricted super tease

Free ChatGPT app for iOS launched by OpenAI

Beware of fake ChatGPT apps! They can scam you out of money; delete if you downloaded any