Get your free personalized podcast brief

We scan new podcasts and send you the top 5 insights daily.

Dane Knecht describes Cloudflare OS as an internal platform that lets employees use AI without leaking data, while teaching agents company-specific practices. It is also a reference architecture for building a company OS on Cloudflare's developer platform, so security is built in rather than added later.

Related Insights

An enterprise-grade AI agent is more than just an LLM; it's a set of instructions governed by a dedicated "trust layer." This layer is critical as it prevents third-party models from learning from proprietary data, ensures customer privacy, and enforces brand guidelines, making it safe to deploy AI with sensitive information.

Standard agent security (allow/disallow tools) is too blunt. Databricks' Omnigens uses stateful, "contextual policies" that track an agent's session history. For example, it might block publishing to a website *if* the agent previously accessed a confidential document in the same session, preventing data leaks.

The future of work involves potentially millions of AI agents operating within a company. This requires a new governance layer, including agent inventories, inspectable reasoning traces, identity management, and sandboxed execution environments to maintain security and control.

For agent frameworks like OpenClaw, the key value isn't just technical features (which are replicable) but establishing a trustworthy, community-governed ecosystem. Users entrust agents with sensitive data, making security and a transparent foundation the critical differentiating factor.

To ensure AI agents are trustworthy and can work together safely, Dreamer's architecture includes a central "Sidekick" that acts as a kernel. It manages permissions and communication between agents, preventing uncontrolled data access and ensuring actions align with user intent, much like a computer's operating system.

For security-conscious organizations, using external LLMs to process confidential data poses inherent risks. Building a walled-off, in-house LLM provides a secure alternative for internal knowledge management and AI tooling, as AvePoint did with its "Chat AVPT."

A key barrier to enterprise AI adoption is security and control. AWS's Bedrock Managed Agents provides each agent with its own dedicated compute environment and unique identity. This allows security teams to create specific governance policies for each agent, balancing enablement with necessary guardrails.

As autonomous agents become prevalent, they'll need a sandboxed environment to access, store, and collaborate on enterprise data. This core infrastructure must manage permissions, security, and governance, creating a new market opportunity for platforms that can serve as this trusted container.

The CEO of WorkOS describes AI agents as 'crazy hyperactive interns' that can access all systems and wreak havoc at machine speed. This makes agent-specific security—focusing on authentication, permissions, and safeguards against prompt injection—a massive and urgent challenge for the industry.

While public discourse on AI safety focuses on existential risk, for enterprises, safety means protecting proprietary knowledge ("alpha"). True enterprise AI safety is achieved by owning the compute, models, and data stack, preventing model providers from stealing trade secrets and customer data.

Cloudflare OS encodes company principles for safe agent use | RiffOn