Bookmarks

    Nothing saved yet. Use the bookmark on any article. Bookmarks are kept on this device only: they stay private and work without an account.

    Settings

    Appearance

    Reading mode

    Font size

    Content width

    Line spacing

    Keyboard shortcuts

    /
    Search
    j k
    Next / previous story
    ← →
    Previous / next section (columns)
    b
    Bookmarks
    s
    Settings
    r
    Columns / scroll
    t
    Light / dark

    Saved in this browser only.

    Ping2Prompt
    Subscribe to the Weekly Brief Sign in
    Green cascading characters and symbols of computer code streaming down a dark screen.
    Photo by Markus Spiske on Unsplash

    ARTEX AI pentest framework used in real attacks; no patch, so harden credentials

    Update (2026-11-19): This advisory has changed: Amazon says the vendor's server-side liveness verdict rejected the deepfake in every attempt the agents recorded., according to aws.amazon.com/blogs/security/amazon-threat-intelligence-observes-threa….

    Update (2026-11-19): Workaround: Barracuda told a minority of XDR customers to revoke and recreate integration credentials; it did not say how many customers were advised. (previously Barracuda says it directly told a minority of its XDR customers to revoke and recreate integration credentials, without giving a number.), according to techcrunch.com/2026/11/16/fbi-says-artex-ai-open-source-hacking-tool-us….

    What's fixed

    Nothing is patched here: ARTEX is an open-source AI penetration testing framework that attackers are using, not a vulnerable product, so there is no fixed version. Amazon and the FBI (as reported by TechCrunch) describe real intrusions, and the sources give no CVE, severity score or affected-version list. Amazon says the biggest campaign stole AI model API keys from exposed LLM gateways such as One API and New API.

    The FBI says some intrusions went from targeting to compromise in under three hours, helped by weak defenses such as missing multi-factor authentication. The creator took ARTEX down on October 8, but Cloudflare counts over 12,000 repositories now.

    What to do

    Put any AI gateway behind authentication and rotate the upstream provider keys it holds. Revoke credentials of former vendors and staff: Barracuda's breach began with one unrevoked vendor credential. Cloudflare's vendor guidance adds rate limits, bot defense and identity-gated access for high-risk apps. Action: revoke stale credentials, require MFA and close exposed AI gateways.

    Sources

    Image: Photo by Markus Spiske on Unsplash


    This story was produced with AI assistance from the sources listed above and fact-checked against them before publication. Read our editorial methodology and the corrections log. Spotted an error? Email [email protected].

    Continue

    Share on Bluesky · Mastodon · Email ·

    Discussion