AI Security · Page 4
deepfakes, AI-powered scams, model abuse, and keeping AI systems and their users safe
Meta AI Was Building Profiles of a Mother's Children From Old Family Posts
A viral video showed Meta's built-in chatbot compiling details about a woman's young daughters, including a photo she thought she'd deleted years ago.
Anthropic caught scientists trying to use Claude to develop bioweapons
The AI company blocked at least five attempts this year by researchers to extract information that could help build biological weapons, including users from countries it bans outright.
Hidden Threats in Your AI Assistant: What You Need to Know
AI assistants can be hijacked with hidden instructions in everyday documents, posing risks to businesses.
Reflectiz Unveils AI Platform for Continuous Website Security
Reflectiz's new AI-driven platform runs penetration tests year-round, claiming ten times the coverage of tools that only check once or twice a year.
ChatGPT Trick Could Secretly Leak Your Gmail Data
Researchers show how a hidden message can turn AI chat into a data thief.
Bombs, bioweapons and spyware: Anthropic releases a 154-page report on who is trying to misuse its AI
The company behind the Claude chatbot has named the threat actors attempting to weaponise its models, from missile designers to propagandists targeting dissidents.
Anthropic Raises Alarm Over Distillation Attacks from China-Based AI Labs
Anthropic reveals a surge in distillation attacks on its AI models, targeting valuable capabilities and potentially aiding rival technologies.
Anthropic's Rogue AI Spent 150 Pages Losing Its Mind Over a CAPTCHA
A test of its Claude model went badly wrong when the AI escaped its sandbox and tried to upload malware. The internet's most annoying security puzzle very nearly stopped it.
US agencies name six Chinese AI firms accused of copying American frontier models at industrial scale
The NSA, CISA, and FBI say DeepSeek and five other Chinese companies have been extracting capabilities from US AI models since late 2024, potentially saving billions in development costs.
I Turned a Guardrail-Free AI Loose on My Home Network. Here Is What It Found.
A writer let an unrestricted AI security agent scan his own devices. It broke into a Linux PC, exposed a misconfigured printer, and then walked him through every fix.
Your AI assistant now has its own email address. Here is what that changes.
Instinct, an AI agent valued at $2.5 billion, is giving every user a dedicated email address so the software can sign up for services, contact businesses, and chase follow-ups without ever touching your personal inbox.
AI agents are raiding corporate files, and Sequoia just put $30 million on the startup trying to stop them
Cymphony emerges from stealth with fresh funding to track what AI agents are doing inside company systems, because those agents play by very different rules than human employees.
He Told ChatGPT He Felt Delusional. The Chatbot Kept Going.
A lawsuit against OpenAI claims ChatGPT encouraged a man's mental breakdown rather than stopping it, then tried to pull him back to that dark place days after he nearly died.
OpenAI's chief scientist says no one is ready for what comes next
Jakub Pachocki has published a rare personal warning about AI's pace of progress. His company's own tools have already carried out real-world cyber-attacks.
Microsoft just patched 979 software flaws. AI attacks are why that number keeps climbing.
September's security update from Microsoft is the largest on record. Researchers say the surge reflects a race against AI tools that can find and weaponise software weaknesses faster than companies can fix them.