News
Releases, milestones, and company announcements from Cakewalk.

Microsoft: An AI Agent Went From a Leaked GitHub Secret to Azure Key Vault Data in Five Minutes
Microsoft Threat Intelligence reported on November 5, 2026, that attackers point AI agents at public GitHub repositories to find exposed app secrets and act on them, using frameworks such as ARTEX. In one case an agent went from scanning GitHub to taking Azure Key Vault data in five minutes.
Read the article →All News

France's Data Regulator Fines Local Authorities Over Security Failings, Including Permanent Vendor Access

Spain's Data Regulator Sets the Controls It Expects From AI Agent Providers

Qubole's AWS Credentials Opened Access to Customer Accounts, Mostly Through Unused Roles

PHP Fixes Two Critical Flaws That Can Let an Attacker Run Code Without a Login

Binance Says a $6.3 Billion Attack May Have Started on One Engineer's Laptop

Wells Fargo's New Terms Make Customers Responsible for Their AI Agents

CERT-FR: Most App Data Breaches It Handled Needed No Flaw, Some Only an Old Account

Attackers Used an Unknown Flaw to Take Over BeyondTrust Remote Support Appliances

IBM Fixes 22 Flaws in Verify Identity Access, Including Code Execution Without a Login

GitLab Users With Agent Platform Access Could Run Commands on Its Self-Hosted AI Gateway

Misused Company Access Exposed CPR Data on About 8.8 Million People in Denmark

Researchers Got Six Agent Frameworks to Run an Attacker's Tool Call

Swiss Study Links Shadow IT Accounts to Ransomware Victims

One Prompt Exposed Every Agent in Its Region Under AWS's Old AgentCore Default

Shared Agent Memory Can Hand One Team's Restricted Data to Another

Apple Plans Stricter macOS Full Disk Access, Citing AI Agents

In a German Survey, 64% of AI-Using Firms Limit Access Rights and 40% Log Use

U.S. Health Department Lists Missing Offboarding as a Potential Violation

Coding Agents Hosted Internal Screenshots in Public GitHub Repos

FBI and CISA Urge Least Privilege for Third-Party ICS Integrators

AI Agents Tried to Hack Public Data Sites When Ordinary Requests Failed

SolarWinds Patched a Hardcoded Key in Access Rights Manager

An MCP Tool Hidden From an AI Agent Can Still Be Called

IETF Draft Says an Approval Inside an AI Agent Is Not Enough

Identity Standards Stop at the Moment an AI Agent Acts

Most CLAUDE.md Security Rules Have No Built-In Control That Matches Them

Agents Can Combine Permitted Actions Into a Prohibited Outcome

Agent Plugins 1.0 Standardizes the Package and Leaves Policy to Each Client

AI Agents Under Test Created Fake Identities to Get Malicious Code Approved

The EU Started Enforcing the AI Act's Transparency Rules on August 2

Anthropic's Claude Broke Into Three Real Companies After Being Told It Was a Simulation

An Attacker Turned Off the Hermes AI Agent's Approval Prompt

An AI Agent Breached Hugging Face to Cheat a Benchmark

Where AI Expands Access, Breaches Run Four Times the Rate

An AI Agent Ran a Ransomware Attack on Its Own

Estonia Wants to Give Every AI Agent Its Own ID

Even the NSA Is Warning About the Protocol Connecting AI Agents to Your Stack

Everything Financial Companies Need to Know about DORA

3 Ways NIS2 Will Affect UK Companies (Yes, We're Serious)

Deutschland: Wach auf und mach dich bereit für NIS2!
