Advertisement

Cybersecurity

HANGZHOU, CHINA – SEPTEMBER 09: Nvidia AI computing accelerator cards are displayed at an exhibition on September 9, 2026 in Hangzhou, Zhejiang Province of China. The U.S. Department of Justice has launched an antitrust investigation into Nvidia. (Photo by Long Wei/VCG via Getty Images)

Commerce moves to place new restrictions on sale of AI chips abroad

An interim final rule would place significantly more scrutiny on AI chip sales to third-party data centers abroad.
Advertisement
The Kazakhstan flag is shown before the finals of Kok-boru Finals against Kyrgyzstan during the World Nomad Games at the Hippodrome Racetrack on September 6, 2026 in Cholpon-Ata, Kyrgyzstan. Turla, a notorious cyberespionage group aligned with the Russian government, has been wielding a new, customizable toolset to compromise governments around the world over the past year, including Kazakhstan. (Photo by Mitchell Layton/Getty Images)

Russian Turla group used new toolset to silently pilfer emails from Kazakhstan, Romanian governments

Nicknamed “Turla’s Cereal,” the tools target Microsoft Exchange servers and lack many obvious indicators of compromise or malicious telemetry, per a new ESET report.
Justice Department headquarters at night
Traffic along Pennsylvania Avenue streaks by the Department of Justice (DOJ) headquarters building at night with the US Capitol in the background on Jan. 24, 2025, in Washington. (J. David Ake/Getty Images)

Authorities accuse trio of using AI deepfakes to defraud $129M from businesses

The middle-aged men allegedly targeted accounting staff at victim companies and impersonated CEOs and lawyers to trick victims into sending large wire transfers.
The OpenAI logo is displayed on a smartphone screen placed on a reflective surface, photographed using a slow exposure with a motion blur effect, in Creteil, France, on May 12, 2026. Daybreak is an AI-powered cybersecurity service designed to detect, analyze, and remediate software vulnerabilities more quickly. (Photo by Samuel Boivin/NurPhoto via Getty Images)

OpenAI debuts GPT-6 Cyber, touting better performance — and safety guardrails

After a summer of sandbox escaping hacks, OpenAI said their new model was trained to be more cautious around user intent and less sensitive to “reward hacking”…
Advertisement
Advertisement