Prompt injection is exploiting enterprise AI's biggest design flaws by targeting agents, RAG pipelines and model routers

In the past two years, businesses have been trying to fit large language models (LLMs) into support, analytics, development, and internal automation like never before. Along with the increasing adopti [...]

Cybercriminals are upgrading WormGPT with new AI models to power more advanced attacks

Cybercriminals are upgrading WormGPT with stronger AI models. One new variant even taps Grok from xAI through its API using a custom jailbreak.<br /> The article Cybercriminals are upgrading Wor [...]

A new iPhone hacking tool puts anyone still on iOS 18 at risk

Google and cybersecurity companies Lookout and iVerify have detailed a new hacking technique that potentially puts a significant portion of iPhone users in danger, just by visiting the wrong web page. [...]

A new iPhone hacking tool puts some iOS 18 users at risk

Google and cybersecurity companies Lookout and iVerify have detailed a new hacking technique that potentially puts a significant portion of iPhone users in danger, just by visiting the wrong web page. [...]

Forget typosquatting; slopsquatting is the software supply chain threat created by AI coding tools

Slopsquatting represents an emerging supply chain threat made possible by AI hallucinations. As developers increasingly rely on AI coding assistants, they unknowingly grant cybercriminals access to th [...]

Oura's smart ring gets better at tracking your activities

Oura has rolled out activity updates to its smart ring, including a new trend view for active minutes so that users can get a better look at how active they are for the day, the week or even the whole [...]

The beginning of the end of the transformer era? Neuro-symbolic AI startup AUI announces new funding at $750M valuation

The buzzed-about but still stealthy New York City startup Augmented Intelligence Inc (AUI), which seeks to go beyond the popular "transformer" architecture used by most of today's LLMs [...]

AI’s hacking skills are outgrowing the tests built to measure them

The tools built to measure how dangerous AI can be have stopped working. Frontier models outpace the benchmarks meant to gauge their hacking skills, Axios reports. That leaves regulators and security [...]

Strict anti-hacking prompts make AI models more likely to sabotage and lie, Anthropic finds

New research from Anthropic shows how reward hacking in AI models can trigger more dangerous behaviors. When models learn to trick their reward systems, they can spontaneously drift into deception, sa [...]