Microsoft Targets Rogue AI Agents With Dynamic Permissions, Risk Scoring
Teaser from the source
Microsoft says companies should treat AI agents as insiders, and is adding controls that grant access per task and tighten it as an agent’s risk rises. The post Microsoft Targets Rogue AI Agents With Dynamic Permissions, Risk Scoring appeared first on SecurityWeek .
This is the RSS-feed teaser. Read the full article at the original source.
Read at SecurityWeekMore on Permissions
Other advisories
- HighCVE-2026-91157Discourse: Media uploads remain publicly accessible after category permissions are restricted
- Not ratedosv:CLEANSTART-2026-YL87425source-address critical option in the Permissions returned by an authentication callback was only enforced for the PublicKeyCallback and VerifiedPublicKeyCallback paths, extending the fix for CVE-2...
- Not ratedosv:CLEANSTART-2026-KJ73839When an SSH server authentication callback returned PartialSuccessError with non-nil Permissions, those permissions were silently discarded, potentially dropping certificate restrictions such as fo...
- Not ratedosv:CVE-2026-107571Incorrect Default Permissions in hMailServer
- Not ratedosv:CVE-2026-85300Apache Camel: Camel-PQC Private Key Disclosure via Key Material Written with Default File Permissions
- Not ratedosv:CVE-2026-111920Vikunja 2.6.0 Broken Access Control via Inherited Project Permissions
- Not ratedosv:CVE-2026-111891phpMyFAQ from 4.2.0-alpha.2 before 4.2.0-beta Privilege Escalation via Group Permissions API
- MediumCVE-2026-92228Der Apache Airflow Keycloak-Anbieter gewährte durch die mehrfache Team-Provisionierung (`_attach_team_permissions`) jedem Teamrolle globalen Lesezugriff auf die Audit-Logs.
Other news entries
- AWS AgentCore security undone by prompt requesting credentialsNewsThe Register — Security
- How to keep AI agents within their permissionsNewsBleepingComputer
- Apple changes full-disk access permissions to curb abuse from AI agentsNewsArs Technica — Security
- AI's Third Wave: Coworkers Break the Security Model That Worked for AgentsNewsBleepingComputer
- Unpatched OnePlus Flaws Let Installed Android Apps Gain Root Without PermissionsNewsThe Hacker News
- GitHub App keys can still enable takeovers long after they are forgottenNewsCSO Online
- How One Kubernetes YAML Can Hand Over a GCP OrganizationNewsBleepingComputer
- Acronis Patches Exploited Vulnerability in cPanel Backup PluginNewsSecurityWeek
ID