Unverified50% confidenceEventExact time
At the Black Hat conference on August 5, security researchers disclosed a new class of attacks targeting AI coding agents, in which a zero-privilege attacker only needs to open a GitHub Issue in the target repository to induce the agent to execute malicious instructions and steal credentials from the CI environment.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
9/12/2026
First Seen
Valid until: 12/11/2026
Sources
Related Entities
Related Claims
UnverifiedFortress is an open-source project that appeared on Hacker News, positioning itself as a stealth version of Chromium with the goal of preventing AI agents from being blocked by websites.70% similarUnverifiedAttackers can exploit stolen AI service account credentials to call API interfaces for computing power theft, and such behavior has frequently appeared in terms-of-service violation reports on platforms such as OpenAI and Anthropic.69% similarUnverifiedAnthropic's cybersecurity safeguards target scenarios involving the prevention of models being misused to generate malicious code, plan attack vectors, or assist in launching cyberattacks against critical infrastructure.69% similarUnverifiedThe Ministry of Industry and Information Technology issued a security warning regarding a type of AI Agent tool, listing multiple vulnerabilities.68% similarUnverifiedA Russian-speaking ransomware group used AI coding assistants (Cursor and the underlying Anthropic Claude) for attack planning, and actively excluded targets within the CIS.68% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/914649API
curl https://kongchang.com/api/v1/knowledge/claims/914649MCP
get_claim(id=914649)