Latest cybersecurity reporting from selected sources.

Zero-click prompt injection can leak data when AI agents meet messaging apps, researchers warn AI agents can shop for you, program for you, and, if you're feeling bold, chat for you in a messaging app. But beware: attackers can use malicious prompts in chat to trick an AI agent into generating a data-leaking URL, which link previews may fetch automatically.…

Researchers at Radware discovered new prompt injection attacks in ChatGPT agentic features

Zero-Click Vulnerability Let Attackers Weaponize Enterprise AI AssistantGoogle patched a vulnerability in Gemini Enterprise that allowed attackers to steal corporate data through a shared document, calendar invitation or email without any user action or security alerts. No malware was executed, no credentials were phished and no data left through approved channels.

Cybersecurity researchers have disclosed a zero-click flaw in OpenAI ChatGPT's Deep Research agent that could allow an attacker to leak sensitive Gmail inbox data with a single crafted email without any user action

Researchers at Radware found a zero-click flaw in ChatGPT Deep Research agent when connected to Gmail and browsing

Zenity CTO Michael Bargury joins the Black Hat USA 2025 News Desk to discuss research on a dangerous exploit, how generative AI technology has "grown arms and legs" —and what that means for cyber risk.

Cybersecurity researchers have uncovered a jailbreak technique to bypass ethical guardrails erected by OpenAI in its latest large language model (LLM) GPT-5 and produce illicit instructions

A zero-click exploit called EchoLeak reveals how AI assistants like Microsoft 365 Copilot can be manipulated to leak sensitive data without user interaction. This entry breaks down how the attack works, why it matters, and what defenses are available to proactively mitigate this emerging AI-native threat.

Microsoft Patched Flaw Allowing Attackers to Hijack Copilot ResponsesA well-phrased email was all an attacker would have needed to trick Microsoft Copilot into handing over sensitive data until the operating system giant patched the vulnerability. The zero-click prompt injection attack vulnerability received a CVSS severity score of 9.3.

Researchers have found a flaw in Microsoft 365 Copilot that allows the exfiltration of sensitive corporate data with a simple email

A novel attack technique named EchoLeak has been characterized as a "zero-click" artificial intelligence (AI) vulnerability that allows bad actors to exfiltrate sensitive data from Microsoft 365 Copilot's context sans any user interaction

A new attack dubbed 'EchoLeak' is the first known zero-click AI vulnerability that enables attackers to exfiltrate sensitive data from Microsoft 365 Copilot from a user's context without interaction. [...]

If this had been a security drill, someone would’ve said it went too far. But it wasn’t a drill—it was real. The access? Everything looked normal. The tools? Easy to find. The detection? Came too late

Potential Zero-Click Account Takeover Exploit Is Among Identified VulnerabilitiesResearchers at security firm Salt Security have uncovered multiple vulnerabilities in third-party plug-ins used in ChatGPT, including a zero-click account takeover flaw that was triggered when users attempted to install the plug-in using their ChatGPT accounts.

Researchers Created Worm That Can Exfiltrate Data, Spread Spam and Poison AI ModelsResearchers have created a zero-click, self-spreading worm that can steal personal data through applications that use chatbots powered by generative artificial intelligence. Dubbed Morris II, the malware uses a prompt injection attack vector to trick AI-powered email assistant apps.