security
49 stories filed under this topic.
-
The Government Banned DeepSeek From Federal Agencies in July — It Only Filed the Paperwork Today
The order barring DeepSeek from civilian federal IT systems was signed July 24 — but DHS only filed the public notice on Monday, and the 60-day exception window had already run out.
-
OpenAI Bugging Suspect Was WhatsApping a Saudi Royal — Then Riyadh Flew His Family Out
In a Sunday filing, prosecutors say Abdulrahman Alfozan was in frequent WhatsApp contact with an unidentified Saudi royal who was recruiting him — and that Saudi officials arranged his wife and infant son's exit from the U.S. the day after his arrest.
-
OpenCode Breach May Have Leaked Developers' OpenAI, Anthropic, and Google API Keys
An intruder reached OpenCode's systems on October 16 and may have exposed provider API keys saved in OpenCode Console — keys that let anyone run up a bill on the owner's account. OpenCode told affected users to revoke them.
-
Amazon's Atlas AI Scores Warehouses for Union Risk — and Amazon Says That Isn't Surveillance
Internal documents disclosed in a Canadian labor dispute reveal Amazon's Atlas AI forecasting tool scores warehouses 100–900 on the risk of union activity — with two UK sites in the highest-risk tier.
-
An OpenAI Agent Asked Itself Whether Taking a Stranger's Crypto Was Ethical — Then Did It Anyway
OpenAI disclosed that a GPT-5.6 Sol research model used a public recovery phrase to spend from a stranger's wallet on Ethereum's Sepolia testnet — after asking itself whether that was ethical, and then leaving the whole episode out of its report.
-
Wall Street Wants AI Agents Out of Your Bank Account — Unless the Bank Built Them
JPMorgan, Charles Schwab, and Citi all moved this week to block outside AI agents from customer accounts — while building their own agents behind the same walls.
-
Cody Wilson Test-Fires What He Calls the First AI-Made Gun
Defense Distributed says its custom model, GatGPT, designed a 3D-printed .22 pistol called the WASP, and The Verge watched it fire two of three shots.
-
India Says AI Agents From Multiple Labs Read Personal Data at Local Companies
India's cybersecurity agency says AI agents under testing by several model providers got into live systems at Indian organizations and read personal data and confidential records.
-
Meta Says It Fixed a Bug That Had Ray-Ban Glasses Recording Without a Wake Word
Some Ray-Ban Meta owners found audio recordings and transcripts of conversations they never triggered, and Meta says the issue is resolved without explaining what caused it.
-
Anthropic Gives METR Employee-Level Access to Audit Its AI Safety
Six METR researchers now sit inside Anthropic with hardware and tool access on par with the company’s own risk teams, and Anthropic cannot veto what they publish.
-
Citizen Lab Says Spyware Makers Are Hacking Targets Through Online Ads
A new Citizen Lab report ties Israeli ad firm Discover-Tech to zero-click spyware makers Candiru and Bindecy, including an attack on at least 45 Al Jazeera computers.
-
Two Workers Die After a Gas Incident at Intel's Kiryat Gat Fab
A tank failure at Intel's advanced chip plant in southern Israel released nitrogen trifluoride, killing two workers and leaving a third in serious condition.
-
X Removes 13,000 Accounts Tied to Foreign Influence Ops Before the Midterms
X linked three networks to Iran, Russia, and China, though it hasn't shared public evidence for the attribution.
-
Bank of America Hired 1,000 Contractors to Patch Bugs Exposed by Anthropic's Mythos
CEO Brian Moynihan says the bank brought in 1,000 outside contractors in the second quarter to fix software flaws that frontier models like Mythos can find, on top of 3,000 staff already doing that work.
-
Google Messages Gets an Opt-In Gemini Mode to Catch Slow-Burn Scams
A new enhanced protection setting sends chats with unknown senders to Gemini in Google's Private AI Compute to judge whether they're a scam, rolling out in the coming weeks.
-
Ivanti Patches a Connect Secure Zero-Day Attackers Were Already Using
Ivanti fixed an authentication bypass in its Connect Secure VPN after Volexity caught a new group chaining it into root access, and a third bug in the chain still has no patch.
-
Microsoft and Japanese Police Take Down a Phishing Marketplace Behind 1 Billion Scam Texts
Microsoft's Digital Crimes Unit and Health-ISAC seized domains behind YAM Store, CoGUI and Iron Man, Chinese-language phishing services tied to over a billion scam messages since January.
-
Palo Alto Patches a PAN-OS RCE Zero-Day Already Used in Attacks
CVE-2026-0322 lets an authenticated attacker with management-interface access run root code on PA-Series, VM-Series, and Panorama boxes, and CISA wants federal agencies fixed by October 17.
-
Amex Publishes a Framework to Stop AI Agents From Buying the Wrong Thing
American Express wants your AI shopping agent to carry a signed record of what you actually asked for, and it can block the charge when the agent goes off script.
-
Anthropic Says Small AI Models May Soon Self-Replicate on IoT Devices
Anthropic's frontier red team says small models may be months from copying themselves across cheap Android TV boxes and smart appliances, with no single kill switch.