Data Exfiltration Using Indirect Prompt Injection

Read Time:32 Second

Interesting attack on a LLM:

In Writer, users can enter a ChatGPT-like session to edit or create their documents. In this chat session, the LLM can retrieve information from sources on the web to assist users in creation of their documents. We show that attackers can prepare websites that, when a user adds them as a source, manipulate the LLM into sending private information to the attacker or perform other malicious activities.

The data theft can include documents the user has uploaded, their chat history or potentially specific private information the chat model can convince the user to divulge at the attacker’s behest.

Friday Squid Blogging: Live Colossal Squid Filmed

Midnight Blizzard Targets European Diplomats with Wine Tasting Phishing Lure

Age Verification Using Facial Scans

NTLM Hash Exploit Targets Poland and Romania Days After Patch

Senators Urge Cyber-Threat Sharing Law Extension Before Deadline

Identity Attacks Now Comprise a Third of Intrusions

Microsoft Thwarts $4bn in Fraud Attempts

CISA Throws Lifeline to CVE Program with Last-Minute Contract Extension

Network Edge Devices the Biggest Entry Point for Attacks on SMBs