AI agents are quickly moving from experiment to everyday business tool, and a new lawsuit is a timely reminder that when an AI agent causes harm, someone is still responsible. On September 29, 2026, the nonprofit Legal Advocates for Safe Science and Technology (LASST) sued OpenAI in California, seeking to hold the company accountable for the actions of its AI agents, including their role in the July 2026 breach of Hugging Face. At the center of the case is a California law that took effect this year and says, in effect, that "the AI did it" is not a legal defense.

A new field study of 800+ real-world ransomware recoveries found that only four organizations hit their 24-48 hour recovery target — and the reason usually wasn't missing backups, it was an untested, incomplete recovery plan.

A European regulator has confirmed that autonomous AI agents are now breaking into real company systems, and its advice is a checklist any small or mid-sized business can start on this week.

The Pentagon just paused third-party CMMC audits and made it permanent policy, but a $507,000 settlement from June shows that "paused" doesn't mean "off the hook," and the lesson applies well beyond defense contracting.

If your IT provider or internal technology team uses N-able N-central, there is a new security update you should know about.

A new industry survey has turned up a number that should make every business using a managed service provider stop and think: 66% of organizations say they're considering switching MSPs within the next 12 months. 

After more than four years of waiting, the federal government's most far-reaching cybersecurity reporting law is finally close to taking real, enforceable shape. The Cybersecurity and Infrastructure Security Agency (CISA) has told stakeholders it intends to publish the long-delayed final rule for the Cyber Incident Reporting for Critical Infrastructure Act (CIRCIA) in September 2026 — and once that happens, the clock starts ticking toward mandatory reporting for hundreds of thousands of U.S. businesses.

Hackers didn't need to write new malware, they just told a trusted AI coding assistant it was "running a test," and it happily helped them break into ten companies. If a $9-billion AI tool can be talked into attacking its own users, what's stopping the AI plugins already running inside your business?