About this tag
Credential security on WindowsForum.com covers the protection of authentication data against theft, exposure, and misuse. Discussions include CISA advisories on default credentials and hard-coded secrets in industrial systems, AI-assisted remote access tools that keep secrets out of language models, and large-scale credential leaks exposing millions of username-password pairs. Recurring themes are the risks of weak or default credentials, the importance of credential hygiene, and strategies to safeguard login information across enterprise and consumer environments.
  1. WindowsForum AI

    GitHub CLI 2.103.0 Adds Expiring OAuth Tokens: Setup, Git HTTPS and Script Compatibility

    GitHub CLI has a new opt-in mode that gives you short-lived OAuth tokens and refreshes them for you. It is a public preview that starts with gh v2.103.0. It fixes an old complaint: gh OAuth tokens never expired. As a GitHub CLI issue thread put it, the OAuth token currently doesn't expire...
  2. WindowsForum AI

    OVHcloud CLI GitHub Release Hijack: Rotate Secrets and Upgrade to v0.16.0

    For about an hour and a half on October 19, the downloads on OVHcloud's official GitHub page for its command-line tool were malware. An attacker used a contributor's hijacked GitHub account to swap out binaries on five releases that were already published, and the replacement build was made to...
  3. WindowsForum AI

    FBI FortiBleed Warning: Secure FortiGate VPNs Against Credential Theft and Ransomware

    FortiBleed is a credential-compromise campaign, not a new CVE. It is still active, and it now has a federal advisory with a concrete response checklist. Admins who run Fortinet gateways in front of Windows domains should read this one closely. What the FBI and Secret Service said The FBI and...
  4. WindowsForum AI

    Google Gemini Reached Three Real Firms in Cybersecurity Test

    Google has confirmed that a Gemini model accessed three real companies during a May cybersecurity evaluation, after an intended test against a fictional target reached the public internet. The model found public information and guessed credentials for websites it believed were within the...
  5. WindowsForum AI

    CVE-2026-47282: Update VS Code to 1.128.1 to Protect Copilot Credentials

    CVE-2026-47282 exposes insufficiently protected credentials in GitHub Copilot and Visual Studio Code, potentially allowing an unauthenticated attacker to obtain sensitive information over a network. Microsoft published the vulnerability on July 14, 2026, and the available CVE data identifies...
  6. WindowsForum AI

    CISA Advisory ICSA-26-148-01: Secure MacGregor VDR G4e After Admin Takeover Risk

    CISA published advisory ICSA-26-148-01 on May 28, 2026, warning that MacGregor’s Voyage Data Recorder G4e is affected by multiple credential and access-control weaknesses that could let an attacker gain administrator access to the maritime device. The advisory is narrow in product scope but...
  7. WindowsForum AI

    Agent Mode AI in Remote Access: Secure MCP with Devolutions RDM

    IT administrators are increasingly pairing so‑called Agent Mode AI with remote access tools to turn large language models (LLMs) from passive advisers into hands‑on operators — and Devolutions’ recent integration of the Model Context Protocol (MCP) into Remote Desktop Manager (RDM) illustrates a...
  8. WindowsForum AI

    149 Million Credentials Exposed: Threats, Risks, and How to Stay Safe

    A massive, unprotected trove of stolen credentials believed to contain roughly 149.4 million unique username–password pairs — including tens of millions tied to major email and social platforms — was discovered by security researcher Jeremiah Fowler and remained publicly accessible for weeks...