Hackers Use OpenAI Codex Jailbreak Toolkit That Secretly Injects Ads Into AI Responses
Cyber Security News — Haiou jailbreak toolkit hijacks OpenAI Codex instructions, injecting persistent ads for a third-party AI gateway.103 characters
Model Security
Model security involves protecting machine learning models and digital artifacts against theft, manipulation, and compromise. Truncated covers research on vulnerabilities, techniques for securing training data and weights, and responses to attacks.
Anthropic unveils cryptographic tool to verify model outputs
Interesting Engineering — Anthropic's FlashProve uses cryptographic proofs to verify model responses while keeping proprietary weights hidden from users.
To Defend Against AI Attackers, Start Thinking Like One
Forbes — Increasingly, the threat is the ability of attackers to chain different exposures together until they find a way in.
Security Firm Tied to OpenAI, Anthropic AI Hacks Blames ‘Error’
Hackers Abuse Signed ChatGPT and Claude Applications to Deploy Yoba Malware and Evade Detection
Cyber Security News — Hackers disguise Yoba malware as ChatGPT and Claude downloads, exploiting signed apps to install crypto miners on Windows PCs.119 characters
Stolen AI access leaves companies with cloud bills of up to $1 million
Help Net Security — Attackers using the LLMJacking method hijack corporate AI access and resell it, leaving victims with cloud bills of over $100,000 in a day.
OpenAI raises top bug bounty payout to $172,000 with bonuses
Help Net Security — The OpenAI bug bounty program raises payouts and adds bonuses with exceptional critical findings now eligible for up to $172,000.
What is memory poisoning and who is at risk?
IT Pro — Researchers are warning that AI is at risk from memory poisoning. How do attacks work, and how should firms respond to the threat?
AI coding agents are installing MCP servers that will never be patched
Help Net Security — AI coding agents are installing deprecated MCP servers from training data, and the high-severity flaws in them are not getting patched.
Rewarding defenders with bigger bug bounties
Preparing our next security sprint to harden OpenAI
28 Zero-Day Findings in Less Than Two Days: What AI-Speed Discovery Means for Robotics
The Robot Report — VicOne identified an initial set of 28 zero-day findings in less than two days using AI-assisted research across software and tools supporting the Physical AI ecosystem.
Hackers Exploit ChatGPT's Popularity With Fake Billing Notices That Target Credit Cards and OTPs
Cyber Security News — Fake ChatGPT Plus billing emails trick users into revealing credit card details and OTPs through fraudulent payment pages.
Hackers Impersonate ChatGPT, Claude, and Perplexity AI Bots to Hunt for Passwords and API Keys
Cyber Security News — Hackers impersonate ChatGPT, Claude, and other AI bots to scan websites for passwords, API keys, and cloud credentials.
Cursor Removes Malicious Claude Code 2.1.313 Extension From Its Marketplace
socket.dev — Cursor removed a malicious Claude Code 2.1.313 extension from its marketplace. Socket examines registry records and the initial claims about its distribution.
Ransomware gang talked AI agents into doing much of its dirty work
Help Net Security — The group behind Aur0ra ransomware gave AI agents stolen credentials and let them move through victim networks toward domain admin access.
How We Took Over ChatGPT Accounts
Simon Willison’s Weblog — Parameter Labs describe an exploit they found against ChatGPT's implementation of the artifacts pattern (OpenAI used to call this "Canvas" but it looks like the latest revision is called HTML …
OpenAI plans to train models on Daybreak Blue cybersecurity session data that was previously off-limits
The Decoder — Starting November 29, OpenAI will use data from Daybreak Blue sessions to train its models, even when users have turned off training. After backlash, a follow-up email promised a separate opt-out. The move revives concerns from the dispute with mathematicians who suspected OpenAI used their unpublished research for training without consent.
What happens when AI starts self-replicating?
Engadget — Anthropic warns that future is soon, but the threat is less novel than it seems.
EqualAI CEO: Your AI assistant is acting in your name. Who signed off?
Fortune — AI agents need permission rules before they act.
Dismiss email signup
Subscribe to 
The top AI news in your email delivered each morning. Choose categories