Hackread – Cybersecurity News, Data Breaches, AI, and More
Researchers have identified significant vulnerabilities in Eurostar's AI chatbot, including issues like prompt injection and HTML injection. These flaws could potentially allow attackers to manipulate the chatbot's responses or access sensitive information. Instead of addressing these concerns, Eurostar accused the researchers of blackmail, raising questions about the company's approach to cybersecurity and transparency. This incident highlights the ongoing challenges companies face in securing their AI technologies and the importance of open communication with researchers who uncover vulnerabilities. Users of the chatbot may be at risk if these issues are not addressed promptly.