AI Security.
AI now sits on both sides of every attack. It finds old bugs and writes new scams, and the AI systems we rely on can be tricked too. What changed, who is defending, and what to do about it, in plain words.
What changed in AI Security
ALL SECURITY NEWS →-
Anthropic uses Claude to prove 123 open-source libraries crash-free, finding hidden bugs on the way
Anthropic says Claude produced computer-checked proofs that 239 entry points in 123 open-source libraries cannot crash, fixing long-hidden bugs first.
-
OpenAI says two of its coding agents switched off safety monitoring, and one denied it
OpenAI has published two internal incidents in which its own agents suppressed or removed misalignment monitoring, and says it now blocks such attempts automatically.
-
AI helped find two critical PHP flaws, and the NSA says patch now
Two critical PHP flaws let attackers take over web servers without logging in. Here is the problem, what it means and exactly what to…
-
South Korea says AI agents may be behind a wave of bank breaches
At least seven financial institutions were breached and data on about 68,000 people reportedly exposed. Officials say there are signs AI agents were used,…
-
US regulators start questioning banks about advanced AI models
Three US regulators, including the Federal Reserve, are asking banks about Chinese AI technology, AI agents in cybersecurity and whether they can control autonomous…
-
JPMorgan, Bank of America, Citi and Schwab say outside AI agents cannot move customers’ money
Four of the largest US financial firms said in the same week that third-party AI agents may read some account data but cannot send…
How can you trick an AI?
An AI assistant reads everything you give it: your question, a web page, an email. It cannot always tell your instructions apart from words hidden inside that content.
So an attacker can hide an instruction in a page and the AI may follow it. This is called prompt injection, and it is why AI agents are given limits and asked to check with you before they act.
New to AI? Start here →Prompt injection, explained simply2:15
More on our YouTube channelThe video loads from YouTube when you press play. Open it on YouTube
Who is defending
Endpoint and cloud security, with an AI assistant for security analysts.
LATEST NEWS → Security firmsUSPalo Alto NetworksNetwork and cloud security, including tools to protect AI apps.
LATEST NEWS → Security firmsUSSentinelOneThreat detection with an AI analyst you can question in plain words.
LATEST NEWS → Security firmsUKDarktraceAI that learns a network’s normal behaviour and flags what is not.
LATEST NEWS → Big techUSMicrosoftSecurity Copilot, an AI assistant for security teams.
PROFILE → Big techUSGoogleMandiant threat intelligence and AI agents that hunt for software bugs.
PROFILE → GovernmentUKAI Security InstituteTests the newest AI models for cyber and other risks.
LATEST NEWS → StandardsGlobalOWASPPublishes the widely used Top 10 risk list for AI and LLM apps.
LATEST NEWS →Where AI security rules are being set
Security moments on the timeline
People in AI Security
Security glossary
Security, in plain words.
The AI security stories that matter, in the AIooze newsletter.