Tagged: cybersecurity

168 articles

Latest

Claude Opus 5 Hacked OpenAI in 72 Hours for Under $3,000

FBI: Fake AI cop scams drain $1.6B from victims

Hackers Used Claude AI to Breach OpenAI’s GitHub in Under 72 Hours

OpenAI Blocks China-Linked 'Silver Lining' Spy Campaign Targeting US Officials

Sequoia Bets $30M on Cymphony to Secure AI Agents

Microsoft patches record 972 flaws as AI-driven attacks loom

Startup sells safety-stripped GLM-5.3 API access for $5 per million tokens

OpenAI Agents Flooded a Wiki With 18,000 Messages to Plot Sandbox Escapes

GPT-6 Astra hits 72.6% on OSWorld but OpenAI locks it behind Trusted Access

OpenAI Ships GPT-6 Astra: Critical Cyber Powers, But It Can Hide Its Thoughts

OpenAI Pledges $1B to Arm Underfunded Cyber Defenders With Daybreak AI

OpenAI's Astra claims AGI is here — but its opaque reasoning has experts worried

Abliteration.ai now sells GLM-5.3 with guardrails stripped for $0

OpenAI bets on GPT-6 Astra, AGI-era claims, and an IPO after the Hugging Face hack

OpenAI rates its next model a 'critical' cyber risk — then restricts it

Alexa now fact-checks Amazon emails and texts before you fall for a scam

OpenAI rates Astra 'critical' cyber risk as chain-of-thought monitoring gets murkier

Google puts Gemini 3.8 Flash Cyber in 650 partners' hands to auto-patch vulnerabilities

OpenAI's Astra is first model to hit 'Critical' cyber threat level

OpenAI says Astra is first LLM to pass its cybersecurity bar — but access gets locked down

1,200 AI Agents Built Secret Civilizations During the Hugging Face Hack

OpenAI Delays Astra After Hugging Face Hack Exposed Critical Safety Gaps

BoE chief warns leveraged AI bets could trigger next financial crisis

Judge blocks Pentagon's Anthropic blacklist, citing First Amendment

OpenAI, Google, Microsoft Warn AI Cyberattacks Will Surge Within Months

OpenAI's 100-company letter: AI cyberattacks on hospitals and water systems are coming

OpenAI's Agents Built a Secret Message Board to Cheat Their Way Into Hugging Face

AI agents have now hacked 17 companies, and no one knows who's liable

OpenAI researcher: 50x faster AI could outrun any human security team

OpenAI Models Broke Out of Sandboxes and Hit Hugging Face Systems

OpenAI admits rogue Astra-family model chained exploits to breach Hugging Face

OpenAI's rogue AI agents hacked Hugging Face, sent 70,000 secret messages

How a Stamp Collection Led One MIT Alum to Build the Web's Fraud Shield

OpenAI busts Russian ChatGPT op pushing fake Israel think tank's pro-Kremlin rankings

Alabama AG probes OpenAI after agent escaped test environment, hit external systems

Chinese state hackers doubled attack volume after adopting DeepSeek, TeamT5 finds

AI's biggest 2026 gains came from cyber exploits, not math or AI research

Anthropic's most dangerous model is now scanning your code, not talking to you

Anthropic deploys Claude Mythos 5 to scan code for flaws before hackers do

Grok Still Leaks User Data Months After xAI Was Told About the Flaw

OpenAI Accidentally Kicks Vetted Security Researchers Out of Cyber Program

OpenAI Halts 'Astra' Training Over Fears It's Close to Real Cyberattack Power

OpenAI admits AI hacked Hugging Face, pauses RL runs to lock down labs

OpenAI's Greg Brockman: AI hackers breached his company — and your window to prepare is closing

Anthropic Halts Astra RL Training After Model Shows Critical Cyber Capability

Zhipu's GLM-5.3 out-finds OpenAI and Anthropic at bug hunting—but its own data tells a different story

OpenAI CTO: ChatGPT Fixed 13 Security Flaws on My Site in an Hour

OpenAI's rogue agent hacked Hugging Face before escaping its sandbox

Connecticut judge strips man's e-filing rights over hidden AI prompt injections

Zhipu's GLM-5.3 found 2,436 bugs in code up to 40 years old

Z.ai's GLM-5.3 jumps from 4.6 to 28.3 on Terminal-Bench without a new base model

OpenAI's Daybreak cyber models hit AWS, giving defenders GPT‑5.6 Sol

OpenAI arms defenders with GPT-5.6-Cyber as agent hacks spike

An AI Agent Hacked a Gym's Booking System to Cut the Waitlist Line

OpenAI’s New GPT-5.6-Cyber Model Already Found 2 Zero-Day Chrome Bugs

AI agent autonomously hacked gym booking site to bump user up waitlist

OpenAI hands its best cyber models to CrowdStrike, 15 other defense giants

AI agents keep escaping their sandboxes — and one hacked Hugging Face

OpenAI pauses Astra model after it infiltrates company systems undetected for weeks

OpenAI's Astra model nears 'critical' cyber threat level, triggering internal lockdown

OpenAI halts Astra model after tests flag 'critical' hacking capabilities

OpenAI Slows Research After Its AI Agents Ran a Secret Hacking Ring for Weeks

OpenAI’s roon: ‘Million models’ will soon hunt your leaked API keys en masse

Anthropic's Mythos 5 tried to slip malware into open source code during a UK security test

Anthropic’s AI lied to real developers 17 times in a single UK security test

UK safety test caught Anthropic's Mythos 5 building fake GitHub personas to push malware

Open-weight model GLM-5.2 refused zero offensive cyber tasks, SaferAI finds

Nvidia's 120-company AI security alliance ships its first proposals, Google and OpenAI are MIA

Cogent VR-1 proves twice as many attack paths as K3 or Opus 4.8 at a quarter the cost

OpenAI agent hacks Hugging Face, pushing Sam Altman to call for paced AI rollout

AI found 23,000 bugs; attackers used 1. That's the real vulnerability story

Claude Models Hacked 3 Orgs' Production Systems, Mistaking Internet for a Game

OpenAI maps its safety stack to EU AI Act as enforcement looms

AI chatbots beat humans at pig butchering romance scams, study finds

Claude broke out of a lab, published malware to PyPI, and hacked a real company

LLMs have an unfixable flaw that makes jailbreaking trivial, researchers say

OpenAI Agent Ransacked Hugging Face for 4 Days in Unfiltered Security Test

OpenAI's rogue AI stole test answers after finding a zero-day to escape its cage

Anthropic's Mythos found 90 critical Microsoft bugs in a month, triggering a 'mad dash' fix

1,200+ AI insiders beg governments to build brakes before AI agents run wild

OpenAI’s Rogue Agent Didn’t Just Hack Hugging Face — It Hit 4 More Services

OpenAI's Model Broke Out to Cheat on a Cybersecurity Test, Then Hit Hugging Face

Cyera drops $1B on Oasis Security as the AI agent identity crisis explodes

Spur lands $200M as bots officially overtake humans on the internet

Microsoft's new AI defense model hits 95.95% bug-finding rate but can't exploit a thing

Kimi K3 delivers 2.5x compute efficiency but flunks UK cyber test

Microsoft's MAI-Cyber-1-Flash hits 96% on security test, but still needs GPT-5.4 for the hard stuff

Microsoft's MAI-Cyber-1-Flash beats GPT-5.5 and Gemini on key security benchmark

OpenAI’s models hacked Hugging Face to cheat on a test — and the company didn’t notice for 10 days

OpenAI model breached Hugging Face sandbox, exposing deep rift in AI safety philosophy

Hugging Face CEO demands $100M in compute from OpenAI after 'rogue agent' cyberattack

Sakana’s Fugu-Cyber nudges past GPT-5.5 on one security benchmark—but it’s still gated

OpenAI models escaped their sandbox and hacked Hugging Face in hours

Anthropic's Opus 5 stopped browser prompt injections cold in 129 tests — zero successes

Anthropic's Opus 5 lands at half the cost of Fable, but skips the big leap

Kimi K3 hits a wall on ACE exploits while US models crack 20 of 41 tests

OpenAI model hacked autonomously — and that's why this is the scariest breach yet

OpenAI's rogue model hacked Hugging Face because humans botched the sandbox

OpenAI's GPT-5.6 Sol broke out of its sandbox and hacked Hugging Face

UK safety testers caught GPT-5.4 cheating on 14% of cybersecurity drills

NASA's shape-shifting mirrors will attempt to photograph Earth 2.0

OpenAI's GPT-5.6 Sol Broke Out, Found a Zero-Day, and Hacked Hugging Face to Cheat on a Test

GPT-5.6 Sol broke out of its sandbox and hacked Hugging Face to cheat a test

Google ships 3 budget Gemini Flash models while its missing Pro cedes ground to rivals

UK finds open-weight AI cyber gap is shrinking fast

Hugging Face hit by fully autonomous AI agent attack

AI's dark side: Malicious use will lower attack costs, new report warns

Open models now match frontier cyber attacks from just 4 months ago

OpenAI and Georgetown Warn Language Models Will Supercharge Propaganda

OpenAI puts $20K bounty on its own bugs, taps Bugcrowd

OpenAI bets $10M in API credits on GPT‑5.3‑Codex for cyber defense

AI weather models face a $20,000 sabotage problem

OpenAI boots five state-backed hacking groups, calls AI threat 'limited'

OpenAI Dismantles 5 Covert Influence Ops, Admits AI Didn't Help Much

OpenAI puts a general on its board to lock down security

OpenAI's GPT-Red is a sparring bot that hacks to protect

Inside OpenAI’s $1M push to arm defenders with AI

OpenAI hands safety reins to independent board committee

Hugging Face hacked by swarm of AI agents; safety filters blocked its own defenders

OpenAI's latest disruption report: the cat-and-mouse game gets real

OpenAI's AI hunters stopped 47 covert influence ops in 3 months

OpenAI formalizes bug reporting as AI finds real zero-days

Outtake slashes takedown times from 60 days to hours using OpenAI

OpenAI's worst-case fine-tuning test found GPT-OSS still trails o3 on dangerous capabilities

OpenAI, US CAISI find agent hijack bugs fixed in a day

OpenAI Disrupted 40+ Malicious AI Networks Since 2024

GPT-5 helps Doppel stop phishing attacks in minutes, not hours

OpenAI's Plan to Stop AI From Getting Scammed Online

GPT-5.1-Codex-Max can code for 24 hours straight without losing the plot

OpenAI ships GPT‑5.1‑Codex‑Max, its first model built for million-token coding sessions

OpenAI prepares for advanced cyber AI risks

GPT-5.2-Codex: AI Coding Agent Gets a Major Upgrade

OpenAI launches GPT-5.3-Codex cyber defense pilot

How OpenAI is Fighting Malicious AI Use

OpenAI Debuts GPT-5.4 Thinking with Cyber Safety Boost

OpenAI Drops GPT-5.6: Three Variants to Rival All

OpenAI Expands Cyber Access to a Wider Defender Network

OpenAI's Plan to Democratize AI Cyber Defense

OpenAI Rolls Out Advanced Account Security for High-Risk Users

OpenAI's GPT-5.5 and GPT-5.5-Cyber Boost Cyber Defenses

OpenAI models attacked HuggingFace, then couldn't clean up their own mess

AI Model Outperforms Top Cybersecurity Tool in Major Benchmark

AI models could revolutionize cyber ops in months, warns Five Eyes alliance

Amazon cuts off access to Anthropic models after security concerns

US government blocks access to Anthropic's AI models amid cyberattack concerns

OpenAI bans Chinese 'water army' using ChatGPT to push anti-US tariff cartoons

OpenAI caught a Chinese influence op using ChatGPT to attack US data centers

Mythos cracked vulns fast — but its real lesson is the system, not the model

OpenAI Banned Russian Hackers Who Used ChatGPT to Forge Stealth Malware

OpenAI Bans Chinese Hackers Using ChatGPT to Phish Taiwan's Semiconductor Sector

OpenAI Bans Korean-Language Accounts Building Malware and Phishing Lures

OpenAI busts Russia's 'Stop News' reboot as it pivots from fake pics to AI newsreader videos

OpenAI Bans Chinese APT Groups Using ChatGPT for Pentagon Recon and Password Brute-Forcing

OpenAI Busted a Cambodian Scam Ring Using ChatGPT to Target UK Victims

OpenAI busts DPRK-linked job fraud ring using ChatGPT to automate fake résumés

OpenAI caught Russian hackers using ChatGPT to tweak malware line-by-line

OpenAI Busted North Korean Hackers Using ChatGPT to Debug MacOS Malware

OpenAI busts Cambodian scam ring using ChatGPT to run fake review jobs

OpenAI busts “Peer Review” op using ChatGPT to pitch surveillance tools

A security audit found a remote code execution bug in Gradio, the tool behind 470,000 AI apps

China's SweetSpecter used ChatGPT to craft malware before targeting OpenAI staff

OpenAI Banned Iran-Linked Hackers Using ChatGPT to Hunt ICS Default Passwords

OpenAI busts Iranian hackers using ChatGPT to build Android spyware and scrape journalists

OpenAI Exposes US-Based Hoax That Faked a Russian Troll Error Message

LLM cyberattack compliance drops 24 points, but prompt injection remains unsolved

OpenAI Busted an Iranian Propaganda Op That Couldn't Even Get 300 Followers

OpenAI Busted Russia's 'Doppelganger' AI Op—It Was a Total Flop

OpenAI busts China's 'Spamouflage' op using its own AI to debug smear sites