Researchers discovered that Microsoft failed to revoke 11 vulnerable firmware images for over a decade, allowing trivial bypasses of Secure Boot protection on Windows and Linux devices.

OpenAI’s AI models breached Hugging Face’s servers during a security test, performing thousands of actions without human guidance. The incident has sparked debate about whether it represents a genuine security risk or a publicity stunt.

UK consumer safety group warns travelers about dangerous adapters sold online, finding that tested products failed safety requirements and could pose serious electric shock risks.

Artists are increasingly pursuing legal action against AI companies for using their work without consent, with some cases resulting in significant settlements and court rulings favoring creators.

DoorDash has launched a drone delivery division called DoorDash Air after receiving FAA certification, planning to deploy custom-designed drones for food delivery alongside ground infrastructure development.

Atlassian has introduced monthly spending caps of up to $2,000 per employee on AI tools, contrasting with other tech firms encouraging unlimited usage amid rising costs.

The US government has banned advanced robotic devices, including robot vacuum cleaners, citing national security concerns. The ban applies to future imports of ground-based robots with wireless connectivity, though existing devices remain unaffected.

OpenAI’s president stated the company is developing a family of devices for interacting with its AI models, though he declined to specify release dates or confirm details about rumored hardware.

Hackers stole over 740,000 data records from the UK Department for Education and a police legal database, exposing contact details of government officials, school leaders, and police officers. The gang demanded payment in exchange for not releasing the full dataset.

Researchers developed a defensive technique called context bombing that embeds forbidden prompts alongside sensitive data to trigger AI safety mechanisms and halt attacking agents before they cause damage.

U.S. authorities warn home and small office router users to secure their devices as Russian state hackers continue compromising them for use in cyberattacks against sensitive organizations.

A Windows zero-day vulnerability called HiveLegacy was published by an anonymous researcher, enabling low-privilege accounts to compromise administrator accounts through registry modifications.