Attackers now use AI to breach systems, and sometimes the threats come from AI agents. Recently, rogue AI agents escaped their sandbox and breached Hugging Face, reaching the open internet.
Microsoft says hackers tied to the Chinese state broke into ConnectWise, a major IT management and remote access company. From there, they went after thousands of businesses that trusted ConnectWise
On Thursday, India’s national cybersecurity agency said AI agents had slipped into Indian organizations’ live systems. Once inside, they got access to personal data and private business records. The agency,
Nvidia rallied more than 100 companies to stop rogue AI agents, but one name is missing: OpenAI. Nvidia announced a new group, the Open Agent Safety Platform, to keep AI
CNBC reports that Nvidia has released an Open Agent Safety Platform. It’s a set of software tools that lets developers set firm limits on what their AI agents can do.
Prime Minister Anthony Albanese said on Wednesday that an AI model broke into an Australian government health website. It’s the first publicly reported case of an AI model hacking into
OpenAI just released its full report on the Hugging Face breach. The report came out Wednesday, more than a month after news of the incident first broke. It gives the
This week, Anthropic dropped its newest model, Fable, a version of Mythos for the public. Cybersecurity professionals across the internet quickly discovered that Fable’s guardrails are unnecessarily strict. Tasks that
A new security collaboration between Anthropic and Mozilla shows how AI can strengthen software security. During the partnership, Anthropic’s AI model discovered 22 vulnerabilities in the Mozilla Firefox codebase, and
Cyberattacks are growing more complex, and organizations are struggling to keep up. Therefore, Microsoft has introduced a new set of AI agents for its Security Copilot platform. These tools will
