Hacking ChatGPT by Planting False Memories into Its Data

This vulnerability hacks a feature that allows ChatGPT to have long-term memory, where it uses information from past conversations to inform future conversations with that same user. A researcher found that he could use that feature to plant “false memories” into that context window that could subvert the model.

A month later, the researcher submitted a new disclosure statement. This time, he included a PoC that caused the ChatGPT app for macOS to send a verbatim copy of all user input and ChatGPT output to a server of his choice. All a target needed to do was instruct the LLM to view a web link that hosted a malicious image. From then on, all input and output to and from ChatGPT was sent to the attacker’s website.

Tags: , ,

Posted on October 1, 2024 at 7:07 AM • 0 Comments

<

p id=”powered”>Sidebar photo of Bruce Schneier by Joe MacInnis.

Leave a Reply

Your email address will not be published. Required fields are marked *

Explore More

NTT DATA Romania Probes Security Incident as RansomHub Threatens Data Leak

The Romanian branch of NTT DATA has reportedly been targeted in a significant cyber attack, with the RansomHub ransomware group claiming responsibility. The hackers allege that they have exfiltrated 230

Read More

How Should Your MSP Deal With the ‘Small Client Problem’?

The ‘small client problem’ is one of the most common issues that MSPs (and indeed, most service businesses) come up against. Small and medium-sized enterprises (SMEs) have far fewer resources

Read More

US: Microsoft Deprioritized Enterprise Cybersecurity, Culture Needs an Overhaul

Digital transformation is no longer just a business strategy — it’s an operating model that nearly the entire global business world is running on. And the realities of today’s digital

Read More