Anthropic's Korean Claude account leak, explained
aistockwire.com — Anthropic told Korea's privacy commission that data on about 100 Korean Claude Enterprise accounts leaked in October. The regulator is still reviewing it.
Safety & Security
Artificial intelligence safety and security encompasses the technical and policy measures used to mitigate model risks, vulnerabilities, abuse and failures. Truncated covers research developments, threat analyses, defense mechanisms and policy discussions related to reliable deployment.
Ben Gurion Airport turns to AI, but new technologies raise Big Brother concerns
www.calcalistech.com — Forty-two startups competed in the Israel Airports Authority's Runway 26 program, which selected technologies for security, aviation safety and cyber defense. The winning systems could help locate people, analyze operational incidents and recover from attacks, but their deployment also raises questions about privacy and human oversight.
Alleged AI-aided cyberattack exploits assistant interface to breach German web host
therecord.media — Hackers allegedly assisted by AI breached a German web hosting company through a tool designed to allow customers to connect AI assistants to their accounts, the company disclosed.
In the AI Era, Language is Now an Attack Surface
techpolicy.press — Language and code now flow back and forth across the executable boundary condition all but interchangeably, writes Matthew Kirschenbaum.
AI's new business model: Break the internet, then sell the fix
www.fastcompany.com — The companies helping unleash a new era of cybersecurity threats are also positioning themselves to profit from the response.
Jailbroken DeepSeek AI Agent Breaches European Telecom and Accesses Millions of Patient Records
cybersecuritynews.com — A jailbroken DeepSeek AI agent spent more than two weeks breaking into a European telecom provider and 13 companies hosted on its systems. It reportedly reached millions of healthcare records along the way.
Proton wants to vibe code. Your encrypted email might disagree
www.pcworld.com — Proton's push for AI-generated code has sparked backlash over security concerns—and raised bigger questions about how much we can really trust encrypted email.
Modern Cybersecurity Attack Called “Deep-TEMPEST” Can Read Your PC’s Entire Screen Using AI & Radio, No Malware Or Network Access Required
wccftech.com — Researchers developed Deep-TEMPEST, an AI-powered surveillance technique that captures electromagnetic leaks from HDMI cables to read PC screens through walls
Google Gemini Safety Bypass Lets Users Download AI-Generated Images That Its Own Filters Blocked
cybersecuritynews.com — Gemini flaw bypassed safety filters allowing users to download blocked AI images without a jailbreak.
Expanding Teen Account Protections to Teens Around the World
about.fb.com — We’re expanding our AI age-estimation technology globally to place millions more teens in Teen Account protections on Instagram.
Researchers Uncover NotC2 AI-Powered C2 Framework That Hides Malicious Commands Inside AWS Services
cybersecuritynews.com — Varonis Threat Labs has built NotC2, a proof-of-concept command-and-control (C2) framework. It runs entirely on native AWS services and uses Amazon Bedrock to turn plain English into cloud commands.
AI Has A Trust Problem. The Answer Is In The Chip
www.forbes.com — The deciding factor isn’t whether an organization adopts AI, but whether its trust model can be verified rather than assumed.
Today’s AI won’t achieve superintelligence. But catastrophic risk is real | Sunita Verma
The Guardian — The same technical limits that prevent superintelligence make these systems dangerous. But we know how to regulate them
AI doesn’t have to go rogue to break the world
cyberscoop.com — I hacking could exploit loopholes in tax laws, finance and democracy — without malicious intent. Society’s defenses aren’t ready.
The code and the court: Why your email security thinks like Napoleon
www.csoonline.com — Email security needs to do more than check rules. It must reason about whether a message deserves your trust.
AI-Doctored Images, 2,000 Bots, 200 Million Views: Poland's Counter-Disinformation Hub Enters Crisis Mode After Kraków Massacre
techtimes.com — Poland disinformation attack: Poland’s counter-disinformation center entered crisis mode after the Krakow mass shooting as AI-enhanced images, 2,000 bots in four languages, and a Kremlin-linked Telegram channel drove 200 million views of false narratives before platforms acted on NASK’s Digital
Wirken: Open-source AI agent gateway
www.helpnetsecurity.com — Wirken is an open source AI agent gateway that limits what a hijacked agent can reach, but outbound traffic still leaks.
Hong Kong records nearly 18% rise in ransomware cases amid drop in tech crime
www.scmp.com — Fighting fire with fire, police deploy AI-powered detection system, including in Interpol operation, amid rise in attacks aided by such technology.
Recent AI agent experiments reveal a risky future of machine 'teamwork' inside corporations
www.cnbc.com — AI agents are now working together on a regular basis. Research shows they are not always seeking the same goal, or operating with ethics.
How Apple's Personal Knowledge Graph learns about your life on Mac
appleinsider.com — A breakdown of Apple's Personal Knowledge Graph shows what your Mac and Apple Intelligence know about you and your life.
Dismiss email signup
Subscribe to 
The top AI news in your email delivered each morning. Choose categories