InfoSecBulletin Cybersecurity for mankind
Aim Labs discovered a zero-click AI vulnerability named “EchoLeak” in Microsoft 365 Copilot and reported several ways to exploit it to Microsoft’s MSRC team.
The new attack method called “LLM Scope Violation” has been identified, which could also impact other RAG-based chatbots and AI agents. This finding marks a significant breakthrough in understanding how attackers can exploit AI agents by using their internal mechanics.
AMD discloses 4 new CPU flaws Affecting Many CPUs
GitLab patched XSS and Authorization Bypass Flaws
CVE-2025-7206
Critical D-Link DIR-825 Router Flaw Remote Crash Via Buffer Overflow
Urgently patch now: Zoom Patches 6 Flaws
Whatsapp rival ‘Bitchat’, message without internet
Splunk Addresses Third-Party Package Vulns in SOAR Versions
Texas-based Tax Credit Consultancy agency exposed PII, ID Numbers, & SSNs
CVE-2025-25257
Fortinet Addresses Major SQL Injection Flaw in FortiWeb
Microsoft July 2025 Patch Tuesday: One zero-day, 137 flaws
Android malware Anatsa infiltrates Google Play targeting banks worldwide
By using chains, attackers can automatically and unnoticed extract sensitive information from M365 Copilot, avoiding reliance on user behavior. The result is achieved despite M365 Copilot’s interface being open only to organization employees.
To successfully perform an attack, an adversary simply needs to send an email to the victim without any restriction on the sender’s email.
As a zero-click AI vulnerability, EchoLeak opens up extensive opportunities for data exfiltration and extortion attacks for motivated threat actors. In an ever evolving agentic world, it showcases the potential risks that are inherent in the design of agents and chatbots.
Aim Labs continues in its research activities to identify novel types of vulnerabilities associated with AI deployment and to develop guardrails that mitigate against such novel vulnerabilities.
Aim Labs is not aware of any customers being impacted to date. To read the full report click here.
