prompt injection attacks

A second research team has demonstrated a prompt injection attack against xAI's Grok, using encrypted commands embedded in webpages to…

Avec's new email app features a swipe-left gesture that auto-extracts deadlines from messages and resurfaces them on the due date,…

Hidden text instructions targeting AI systems have moved from academic papers to legal filings, with a July 2026 Connecticut lawsuit…

Cloudflare launched Kitesurf, a cloud-hosted browser designed specifically for AI agents rather than human users, marking a strategic shift away…

Zenity researchers at Black Hat demonstrated that OpenAI's Atlas browser, despite having the strongest security of tested AI browsers, can…

OpenAI developed GPT-Red, a specialized AI model designed to find vulnerabilities in its own systems through automated red-teaming, but has…

AI features added to products are creating high-risk vulnerabilities that are fixed slower than other issues, according to Cobalt's 2026…

OpenAI launched Lockdown Mode to protect users from prompt injection attacks by disabling live web browsing, image retrieval from the…

Early AI jailbreaks were surprisingly simple, using basic psychological tricks like "ignore all previous instructions" or roleplaying scenarios (e.g., the…

Google reports a rise in malicious AI prompt injection attacks, but notes their overall sophistication remains low, with many indirect…

A researcher executed successful indirect prompt injection attacks against AI agent integrations from Anthropic, Google, and Microsoft, stealing API keys…

A major security liability exists as 65% of deployed AI agents are unused but retain live access to critical systems,…

A hacker exploited a known vulnerability in the Cline AI coding assistant to silently install software on users' computers, demonstrating…

OpenAI has introduced **Lockdown Mode**, an optional security feature for its enterprise and education ChatGPT plans, designed to defend against…

AI integration creates severe security gaps, forcing organizations to choose between falling behind or deploying vulnerable systems that attackers actively…

Moltbot is a novel open-source AI assistant that proactively initiates tasks and integrates with popular messaging apps, but its complex…

The cybersecurity landscape in 2026 will be defined by the weaponization of AI, leading to faster, more adaptive, and harder-to-detect…

OpenAI acknowledges that complete security for its AI-powered Atlas browser may be impossible, highlighting a core tension where the tools'…

Gartner advises businesses to temporarily block AI-powered browsers due to significant security risks, like data loss and exploitation, that currently…

The UK's National Cyber Security Centre warns that prompt injection attacks on large language models (LLMs) may be fundamentally unsolvable,…