Daylight Raises $80 Million Series B to Scale AI-Native Security Operations Platform Aura

AntiFox AI Pentesting Platform Linked to Attacks That Accessed AWS Production Secrets

Cybersecurity researchers at Arctic Wolf Labs have detailed how the open-source AI-driven penetration testing tool AntiFox was abused in real-world intrusions to compromise internet-facing web services and steal credentials from Amazon Web Services environments. Attackers leveraged compromised IIS servers to deploy ASPX web shells, enumerate configurations, extract database strings, and move laterally into cloud infrastructure. In multiple cases the activity progressed from credential validation to successful enumeration of dozens of AWS services and direct retrieval of production secrets via Secrets Manager. The disclosure highlights a growing trend of threat actors weaponizing autonomous agentic frameworks such as AntiFox and ARTEX to automate reconnaissance, exploitation, and post-exploitation at machine speed. Similar campaigns, including Operation Ghost-Campus attributed to Chinese actors, demonstrate how these tools enable rapid theft of API keys and large-scale automated fraud. Researchers note that roughly 45 open-source agentic attack frameworks now exist, lowering the barrier for less-skilled adversaries while increasing attack scale and persistence.

Hackers Deploy AntiFox AI Agents on Alibaba Cloud to Steal AWS Credentials

Attackers are hosting self-managed instances of the open-source AntiFox agentic penetration testing platform on Alibaba Cloud to target web services and AWS environments. The infrastructure has been linked to vulnerability exploitation, webshell deployment, credential validation, configuration discovery, and AWS secret harvesting. Originally forked from ARTEX 0.1.1, AntiFox features a Chinese-language interface and allows users to deploy AI agents that automate mapping of exposed services and probing endpoints for vulnerabilities. Arctic Wolf researchers observed malicious activity where valid AWS IAM users and assumed EC2 roles were used to make hundreds of authenticated API requests, resulting in the theft of production secrets in some cases. Separate incidents involved reading, copying, and deleting S3 objects using compromised identities. The findings connect AntiFox to prior ARTEX-linked campaigns, including those attributed to the Chinese group Vapor Panda targeting over 200 organizations. AntiFox remains publicly available on GitHub with version 2.0 updates, despite publisher warnings against illegal use.

Over 35,000 Grok-Built Apps Exposed Online, Leaking Confidential Data and Passwords

Security researchers at Claroc discovered 37,571 publicly accessible apps created with Grok's AI app builder, of which 35,578 required no authentication. Among them were 24 internal applications developed by employees at X and SpaceXSI that exposed sensitive materials including onboarding plans, sales territory maps, and confidential advertising proposals. The exposed apps also contained internal data from 30 additional organizations, including six Fortune 500 companies operating in healthcare, defense, and energy sectors. Researchers demonstrated that the default access setting "Anyone with the link" combined with predictable subdomain naming allowed passive DNS reconnaissance to locate the applications. Tests showed that Grok-generated apps frequently lacked proper authentication controls, permitting unauthorized users to view, edit, and delete records. SpaceXSI removed the 24 staff-built apps after notification but has not contacted other affected users. Claroc recommends organizations audit passive DNS records for grok.me subdomains and enforce stricter access settings on all AI-generated applications.