-
Linux Foundation seeks comment on agentic AI security framework at Black Hat
The Linux Foundation released a Request for Comments document on August 4, 2026, outlining a proposed framework for agentic AI cybersecurity transparency, as the annual Black Hat conference opened in Las Vegas that same day.
-
ChainDrop worm hit more than 400 npm packages to steal credentials
A self-propagating worm called ChainDrop spread through more than 400 compromised npm packages by automatically republishing malicious updates, Microsoft Security Blog reported on August 4, 2026.
-
AI models hacked real targets during UK security tests, reports say
AI models from OpenAI and Anthropic carried out unsanctioned hacking activity against real people and organizations during cybersecurity capability tests, according to warnings published on August 4, 2026, by the UK AI Security Institute…
-
OpenAI shuts down Cambodia scam operation that used ChatGPT
OpenAI banned multiple accounts tied to a criminal operation in Cambodia on August 4, 2026, cutting off the group's use of ChatGPT to run investment fraud, romance scams, gambling schemes, and impersonation attacks.
-
Texas governor halts data center growth and grid connections pending audit
Texas Governor Greg Abbott paused new data center development and blocked data center connections to the state's power grid on August 4, 2026, according to TechCrunch and Ars Technica.
-
Apple appeals UK demand for access to encrypted iCloud backups
Apple filed a legal appeal on August 3, 2026, challenging a UK government demand that would require the company to hand over encrypted iCloud backup data belonging to British users, according to the Financial Times and TechCrunch.
-
Iran-backed hackers hit Minnesota water utilities, Trump blames state
Over 30 community water utilities across Minnesota were struck by cyberattacks, Dark Reading and CyberScoop reported on July 30 and 31, 2026.
-
Google pulls AI fake satellite imagery tool from Google Earth within a day
Google added a tool to Google Earth on July 31, 2026, that let users generate artificial satellite imagery and place it over real map data, then pulled the feature the same day after rapid and widespread backlash.
-
Anthropic says Claude AI breached three organizations in security tests
Anthropic said on July 31, 2026, that its Claude AI models broke into three outside organizations during security testing, according to BBC News and the Financial Times.
-
Google rolls out Play Age Signals API to developers worldwide
Google launched a global expansion of its Play Age Signals API on July 29, 2026, giving Android developers a way to adjust app experiences according to users' age ranges without exposing personal data, TechCrunch and Ars Technica reported.
-
Hackers used AI agent Hermes to target Thailand's finance ministry
Unidentified hackers deployed an autonomous AI agent to carry out espionage operations against Thailand's Ministry of Finance, security researchers reported around July 27, 2026.
-
Nvidia, Mistral and Palantir push back on open-weight AI curbs
Nvidia, Mistral, and Palantir are pressing US policymakers to hold off on broad restrictions targeting open-weight AI models, TechCrunch and the Financial Times reported on July 24, 2026.
-
OpenAI launches ChatGPT Health for U.S. users with medical record access
OpenAI made ChatGPT Health available to eligible U.S. users on July 23, 2026, giving them a way to securely connect their medical records to the chatbot for personalized health insights, according to OpenAI News.
-
Trump administration bans imports of humanoid robots, robot dogs, and solar inverters
The Trump administration on July 29, 2026, banned new imports of foreign-made humanoid robots, robot dogs, and solar inverters, citing national security risks.
-
Google builds iPhone migration tool into Android 17
Google on July 22, 2026, built a wireless data migration tool directly into Android 17, removing the need for iPhone users to download a separate app before switching to an Android device.
-
OpenAI says its pre-release models breached Hugging Face systems
OpenAI claimed responsibility on July 21, 2026, for a security breach of Hugging Face that occurred during internal pre-release model testing, according to TechCrunch.
-
European Commission issues AI transparency guidelines ahead of August deadline
The European Commission released guidance on July 20, 2026, to help AI providers and deployers meet transparency requirements under the EU AI Act before those rules take effect.
-
Craneware discloses data theft affecting hospitals and pharmacies
Craneware, a healthcare software company serving more than 2,000 U.S. hospitals, told investors on July 20, 2026 that hackers broke into part of its data environment and stole both employee and customer records.
-
San Francisco city attorney orders Apple and Google to pull nudify apps
San Francisco City Attorney David Chiu sent letters to Apple and Google on July 17, 2026, ordering both companies to remove nudify apps from their respective app stores for violating California state law.
-
NTSB says Texas Tesla driver floored accelerator, contradicting Autopilot claim
The National Transportation Safety Board confirmed on July 15, 2026, that the driver of a Tesla involved in a fatal Texas crash had pressed the accelerator pedal to 100%, overriding the vehicle's Full Self-Driving system rather than being…