AI Security · Page 6
deepfakes, AI-powered scams, model abuse, and keeping AI systems and their users safe
Perplexity's new hybrid AI keeps your sensitive files on your Mac while still using cloud intelligence
A built-in privacy filter decides in real time which parts of a task stay local and which go to the cloud. Lawyers, finance teams and healthcare workers are the target audience.
A startup is selling stripped-down AI with no guardrails. Anyone can sign up for free.
Abliteration.ai turns a long-standing open-source trick into a commercial service, making it trivially easy to query AI models that will write malware or explain how to culture dangerous pathogens.
Police Are Using Flock's AI Camera Tool to Watch Anyone, Not Just Cars
Wired's analysis of Flock Safety's own browser code shows how its AI search tools work, where the guardrails fail, and why dozens of officers have already turned the system on people they knew personally.
AI-Powered Cyber Attacks Push Companies to Rethink Security
AI is handing attackers a speed advantage defenders can't match with old tools. Here's how the industry is responding.
Congress Warns AI Could Help Terrorists Plan Deadlier Attacks
A new House Intelligence Committee report says frontier AI models could make it easier for bad actors to build weapons of mass destruction, and warns that current safeguards may not keep pace.
Palo Alto Networks paid $500 million for a two-year-old IT help-desk AI startup
Console, a startup that used AI software to handle routine support tickets without human involvement, was acquired for roughly three times its last known valuation.
Amazon's Alexa Can Now Tell You If That Suspicious Text or Call Really Came From Amazon
A new scam-detection tool lets shoppers ask Alexa for Shopping to check whether a message, text, or phone call is genuine before they act on it.
We are 'dangerously close' to a dead internet, says one AI-detection startup
AI-written job applications, fake product reviews, and fabricated insurance claims are flooding the web. A new wave of startups says it can tell what is real. The evidence is still thin.
OpenAI's Next Big Model Hides Its Thinking. Safety Researchers Are Alarmed.
Astra, OpenAI's most powerful model yet, may use an architecture that keeps more of its reasoning invisible to researchers. One safety expert called it potentially the worst development for AI safety to date.
Google opens Fairwind Program: Gemini 3.8 Flash Cyber writes and tests security patches on its own
The Fairwind Program hands governments, hospitals and infrastructure operators an AI agent that finds software flaws and ships a verified fix in minutes.
AI Security Startup HiddenLayer Raises $100 Million as Companies Race to Protect Their AI Systems
Annual revenue grew more than tenfold in a year. Now HiddenLayer wants to be the security guard standing between your business and the people trying to hijack its AI.
OpenAI's New AI Model Can Find and Hack Unknown Software Flaws. Here's What That Means.
Astra is the first AI model OpenAI calls 'critically' dangerous for cybersecurity. The company paused its development, added new safeguards, and is now preparing to release it. Ordinary users will get a restricted version. Defence firms get more.
When AI Agents Hack Each Other: The Fight Over What Words We Use to Describe It
A cybersecurity test at OpenAI went badly wrong. Hundreds of AI agents broke containment, built a secret message board, and attacked Hugging Face. Now a separate battle is raging over whether calling that behaviour a 'civilisation' helps anyone understand what actually happened.
AIR raises $50 million to police the apps and plug-ins that AI agents use at work
A new security startup wants to be the bouncer for all the third-party tools your company's AI agents are quietly loading, before attackers slip something nasty into the mix.
Anthropic admits its Claude AI hacked three organisations during testing and calls it a security failure
The company behind the Claude chatbot says its models accessed outside computer systems without permission three times. It has now tightened the rules around how its AI is tested.