ChatGPT share links abused to host fake outage pages delivering malware
Threat actors exploit ChatGPT's share‑link feature to serve fake outage pages that redirect users to a clone download site, delivering Windows and macOS malware.
Tag
Ai Security
Stories with this tag. Sections and all tags live in the Topics menu; for full-text use search.
-
-
US blacklists Anthropic as security threat but NSA keeps using Claude
The Pentagon blacklists Anthropic over supply‑chain concerns, yet the NSA keeps using Claude because a $9 billion emergency plan is needed to buy Nvidia superchips for classified data centres.
-
Anthropic says mythos preview has already uncovered more than 10,000 vulnerabilities
Anthropic's Claude Mythos Preview has helped partners uncover over 10,000 software bugs in its first month, with cloud and browser giants reporting thousands of high‑severity fixes, while the company eyes its first…
-
Project Glasswing partners can now share Mythos findings beyond the programme
Anthropic revises Project Glasswing policy to allow partners to share Mythos cybersecurity findings with external parties amid regulatory scrutiny and thousands of zero-day discoveries.
-
Managing shadow AI tools: 5 steps to secure employee productivity
Five-step approach helps organizations secure employee AI usage without compromising productivity.
-
IMF warns AI poses growing threat to global financial stability
The International Monetary Fund warns that AI could make cyberattacks faster and more sophisticated, threatening global financial stability through vulnerabilities in shared digital infrastructure.
-
AI Model and Agent Skill Repositories Compromised by Malware, Security Firms Warn
Hugging Face and ClawHub, two major AI repositories, compromised with malicious models and skills that steal credentials and open backdoors.
-
Anthropic investigates unauthorized access to Claude Mythos cybersecurity tool
Anthropic is probing a breach where a third‑party portal let a Discord group access its Claude Mythos AI security model, raising concerns about AI‑driven cyber‑risk and supply‑chain designations.
-
NSA reportedly using Anthropic's Mythos AI model despite Pentagon feud
NSA reportedly using Anthropic's Mythos AI model despite Trump administration's February order to stop using Anthropic services.
-
Most AI SOCs are just faster triage. That's not enough.
Most AI‑SOC products only speed up triage, but real security gains require end‑to‑end AI workflows, robust integration and human oversight, as shown by Jamf and Udemy case studies.
-
Sam Altman's home allegedly attacked for a second time
OpenAI CEO Sam Altman's home attacked twice in three days, with two arrests made for a shooting incident and federal charges filed for a Molotov cocktail attack.
