Autodesk Partners with Permiso Security to Monitor AI Agents
In brief
- Autodesk has joined forces with Permiso Security to monitor AI agents across its cloud infrastructure and workforce.
- This collaboration addresses the growing need for security in AI-driven environments, where agents operate autonomously and can access sensitive data.
- Permiso’s platform provides visibility into agent activities, including tool calls and data interactions, helping Autodesk ensure compliance and reduce risks.
- The move comes as enterprises increasingly adopt AI agents that make decisions without human oversight.
- However, most organizations lack tools to track these agents effectively.
- Permiso’s solution goes beyond traditional security by focusing on runtime behavior, where agents dynamically interact with systems.
- This approach helps identify over-privileges and policy violations in real-time.
- Looking ahead, this partnership highlights the importance of securing AI agents as they become more prevalent in enterprises.
- By integrating advanced monitoring, Autodesk and Permiso aim to set a standard for responsible AI deployment, ensuring both innovation and security coexist.
Terms in this brief
- Permiso Security
- A company that specializes in monitoring and securing AI agents within enterprise environments. Their platform helps organizations track AI agent activities, ensuring compliance and reducing security risks by focusing on real-time behavior analysis.
Read full story at SiliconANGLE →
More briefs
AI Helps Recover $400k in Bitcoin After 11 Years
A cryptocurrency holder who lost access to their wallet due to a forgotten password was able to recover 5 Bitcoin, worth nearly $400,000. The user had been trying for over a decade after changing the password while high and forgetting it. Using Claude, an AI, they dumped old computer files, which led to discovering a backup wallet file from 2019 and a bug in the password setup. Claude successfully decrypted the wallet, allowing the Bitcoin to be transferred. This breakthrough highlights how AI can assist in complex tech challenges, offering new hope for data recovery efforts.
Microsoft Phases Out Claude Code for Copilot CLI
Microsoft is reducing access to Claude Code, replacing it with GitHub Copilot CLI. Thousands of developers were using Claude Code since December, but its popularity has overshadowed the new Copilot tool. The shift aims to streamline processes and cut costs ahead of the financial year end in June. While some employees may find this transition challenging, especially those without coding experience, Microsoft sees this as a strategic move to focus on its own Copilot product, aligning with its engineering needs and security standards.
AI Legal Tools Gain Traction with Lawyers
Anthropic's Claude AI has launched specialized plugins designed to assist legal professionals. These tools cover areas like contract review, compliance monitoring, and regulatory filings. For instance, the "Vendor Agreement Reviewer" evaluates supplier contracts against specific guidelines, flagging issues for attorney review. The system is built with safeguards, ensuring outputs are labeled as drafts needing lawyer approval. Each tool includes source citations, privilege settings, and jurisdictional notes to help attorneys make informed decisions without replacing their judgment. While these plugins speed up processes, they don't provide legal advice or conclusions-only analysis. Looking ahead, such AI tools could transform how legal teams handle routine tasks, freeing up lawyers for more complex work. As the technology evolves, it promises to enhance efficiency and accuracy in legal workflows while maintaining the necessary professional oversight.
Red Hat Expands Support for Sovereign AI to Address Growing Regulatory Pressures
Red Hat has announced new measures to help organizations maintain control over their AI infrastructure amid tightening global regulations. The company is expanding its support for sovereign clouds, which ensure data and operations stay within specific national or regional borders. This market is growing rapidly at 36% annually. Red Hat executives emphasized that sovereignty isn’t just about compliance but about operational resilience and transparency as governments demand more control over AI systems. They introduced tools like "landing zones" built on OpenShift and Ansible, designed to enforce jurisdictional control from the start. The company also launched a service fulfillment API for provisioning sovereign services and localized software delivery within the EU. These steps aim to give customers greater autonomy and choice in managing their AI infrastructure.
Microsoft Aims to Acquire AI Startups for New Model Launch
Microsoft is exploring acquisitions of AI startups to develop a new model, as it prepares to separate from OpenAI. The company reportedly considered buying Cursor but faced regulatory concerns. Now, it's in talks with Inception, where Microsoft's M12 fund previously invested $50 million. This move aims to strengthen competition against OpenAI and others like Amazon. Microsoft has spent over $100 billion on AI investments and infrastructure, making its scale and enterprise ties a key advantage. The new strategy could redefine the AI landscape as Microsoft seeks to expand its capabilities beyond its current partnership with OpenAI.