The real AI cyber divide starts after model access
Frontier cyber tools should be judged by remediation, affordability and downstream protection, not vulnerability counts or partner prestige.
Read More
Author: Alex Mercer
Google starts Gemini 4 Argon rollout with selected cyber defenders
Gemini 4 Argon is reaching selected Fairwind partners first, while public access, independent testing and details of a claimed healthcare flaw remain pending.
Read More
Apple fixes possibly exploited CoreGraphics flaw with iOS 26.7.1 and iPadOS 26.7.1
iOS 26.7.1 and iPadOS 26.7.1 address CVE-2026-86950, which Apple has linked to possible exploitation in a targeted attack.
Read More
OpenAI paused tool-using work on its most capable models after DNS control failure
OpenAI’s pause covered training, evaluation and tool-using inference for its most capable models, while separate government-site incidents had differing evidence and impacts.
Read More
Apple patches CoreGraphics flaw in Tahoe and Sequoia alongside Golden Gate update
Apple’s three Mac updates serve different purposes, with a CoreGraphics security fix for Tahoe and Sequoia and separate maintenance work for Golden Gate.
Read More
Bitget attributes US$388 million theft to third-party zero-day as DPRK link remains unconfirmed
Bitget has outlined a claimed zero-day attack chain behind the US$388 million theft, while independent forensic findings and the exchange’s DPRK indicators remain unpublished.
Read More
Australian state discloses $809,000 loss tied to third-party cyber incident
A Queensland department reported an $809,000 loss linked to illegal access at a telecommunications supplier, while the financial mechanism and recovery status remain unknown.
Read More
Common Defense says security service went live before formal launch
Common Defense reports millions of screened messages, while live integrations, threat definitions, performance evidence and data handling remain unclear.
Read More
The login page is real. The request can still be a trap
A genuine sign-in page may authenticate the right user for the wrong device, application or action. Anti-phishing strategy must verify the process, not merely the interface.
Read More
The 53-image exposure puts OpenAI’s agent controls under scrutiny
The exposure of 53 ChatGPT-user images raises a broader question: can frontier AI laboratories track and contain research agents once they reach external systems?
Read More