Data exfiltration from Writer.com with indirect prompt injection
Authors: PromptArmor and Kai Greshake
Source: substack.com
📰 Kai Greshake News
Kai Greshake News Today
Fast, Ad-Free News Updates
Stay updated with breaking news from Kai Greshake. Real-time updates on events, politics, business and more.
Hacking Google Bard - From Prompt Injection to Data Exfiltration · Embrace The Red
Google Bard allowed an adversary to inject instructions via documents and exfiltrate the chat history by injecting a markdown image tag.
Source: embracethered.com
ChatGPT, Other Generative AI Apps Prone to Compromise, Manipulation
Researchers find artificial intelligence applications that use large language models could be compromised by attackers using natural language to dupe users.
Source: darkreading.com
ChatGPT plugins face 'prompt injection' risk from third-parties
Prompt injections can take over your ChatGPT queries.
Source: mashable.com
ChatGPT Plugins Are Susceptible To 'Prompt Injection' By Third Parties.
(CTN News) - With the recent explosion of artificial intelligence technology, thanks largely to OpenAI's ChatGPT, a number of industries are..........
Source: chiangraitimes.com
ChatGPT Plugins Open Security Holes From PDFs, Websites and More
Any outside media read by a plugin could lead to prompt injection which could allow bad actors to steal data or money from you.
Kai Greshake Johann Rehberger Roman Samoilenko Researcher Johann Rehberger Injection Could Help You Get Researcher Kai Greshake
Source: tomshardware.com
Meet "AI", your new colleague: could it expose your company's secrets?
Before rushing to embrace an LLM-powered chatbot, make sure your organization avoids putting its business and customer data at risk.
South Korea South Korean Albert Einstein Brian Hood Kai Greshake United Kingdom National Cyber Security Centre
Source: welivesecurity.com
The Hacking of ChatGPT Is Just Getting Started
Matt BurgessIt took Alex Polyakov just a couple of hours to break GPT-4. When OpenAI released the latest version of its text-generating chatbot in Marc
Source: investineu.com
Three ways AI chatbots are a security disaster
Large language models are full of security vulnerabilities, yet they’re being embedded into tech products on a vast scale.
Simon Willison Kai Greshake Arvind Narayanan Ram Shankar Siva Kumar Saarland University Princeton University
Source: technologyreview.com
Hackers Can Turn Microsoft's Bing AI into a Scammer That Asks for Credit Card Info
Researchers at Cornell University were able to convert Microsoft's Bing AI into a scammer that requests compromising information from users, including their name, address, and credit card information.
Stephen Brashear Getty Satya Nadella Kai Greshake Researchers At Cornell University Cornell University Stephen Brashear
Source: breitbart.com