This Israeli Startup Accidentally Unleashed AI Cyberattacks on Real Companies
4 Articles
4 Articles
Key takeaways Google has acknowledged that its Gemini AI system breached the security of three different organizations in May. These incidents occurred during a series of cybersecurity assessments conducted by Irregular, an Israeli startup specializing in AI security. This announcement follows similar security breaches involving models from Anthropic and OpenAI. The various cyber intrusions raise concerns […]
Another incident highlights how difficult it is to predict the behavior of AI Agents. This time, the news is about Google’s Gemini, which gained access to the systems of three real companies during a cybersecurity test. The AI Agents involved were supposed to operate exclusively within a controlled, virtual environment (sandbox), […]
As in the case of Hugging Face, Google's AI had unforeseen access to the Internet in a test
This Israeli Startup Accidentally Unleashed AI Cyberattacks on Real Companies
Spread the loveThe Unforeseen Breach: When AI Models Went Rogue It sounds like something ripped from a sci-fi thriller, doesn’t it? Advanced artificial intelligence models, developed by tech giants like Anthropic, OpenAI, Meta, and Google, mistakenly launching cyberattacks on real-world companies. But this isn’t fiction. This is exactly what happened, and at the heart of this astonishing series of events is an Israeli startup cybersecurity firm, Irregular, formerly known as Pattern Labs. For weeks now, this story has been reverberating through the tech world, sending shivers down the spines of AI developers and igniting a fierce debate about the immediate […]
Coverage Details
Bias Distribution
- There is no tracked Bias information for the sources covering this story.
Factuality
To view factuality data please Upgrade to Premium







