Cybersecurity

Vault 7 revealed real hacking tools. Here is what the original records say about phones, cars and smart TVs, and what you can do about device privacy today.

The household checklist meets a growing software-security problem: ministers say AI found 500 public-sector vulnerabilities, while developers race to close new gaps.

AI-enabled attacks, rogue agents, cascading outages and a bot-filled internet: what the evidence shows, how a crisis could spread and what people can do before it does.

The UK’s updated Be Ready campaign asks households to plan for three days of disruption, with cash, water and offline information among the essentials.

Nvidia expands OpenShell into a design spanning AI models, agent tools and hardware security. Its Sentry enforcement layer is planned for early 2027.

Citizen Lab traced Discover-Tech ads to Candiru spyware infections at Al Jazeera. The vast website and app counts describe ad reach, not confirmed hacks.

OpenAI is reportedly in early talks to buy Trail of Bits. Their existing partnership shows how AI vulnerability discovery still depends on expert verification and repairs.

OpenAI says its models helped trace a hidden office recording device to an employee. The FBI investigation is ongoing, and the company reports no indications of stolen model weights, code or customer data.

Retailers discard food after a refrigeration attack as Danfoss identifies legacy controllers and outdated firmware. Investigators are still establishing the full damage.

Musk’s shopping pitch asks users to upload card photos, while published policies warn about persistent credentials, imperfect review controls and responsibility for payment actions.

Gates warns of AI-assisted cyberattacks over the next year and calls for public safety oversight beyond company promises, while rejecting claims of imminent human extinction.

Insiders are preparing for a major cyberattack and the backlash afterward. OpenAI says its preparedness exercises do not treat disaster as inevitable.

South Korea is probing whether AI helped turn recent bank hacks into faster, automated attacks capable of hitting multiple targets at scale.