Anaconda Acquires Enkrypt AI to Help Enterprises Secure AI Agents at Scale

Anaconda has acquired AI security startup Enkrypt AI, adding security, governance and compliance capabilities to its enterprise AI platform. Financial terms of the deal were not disclosed.

Founded in 2012, Anaconda is an Austin, Texas-based software company that provides a platform for open-source Python, data science and AI development. The company helps enterprises securely build, manage and deploy AI applications, serving millions of users and thousands of organisations worldwide through its AI and package management platform.

According to Anaconda, the acquisition is a result of a growing challenge for chief technology officers (CTOs) and chief information security officers (CISOs)– understanding and governing what AI agents are actually doing once deployed.

Recently, a series of cybersecurity evaluations conducted by the UK's AI Security Institute (AISI) has reignited debate over the risks of increasingly autonomous AI systems.

Many organisations today struggle to determine whether the AI-powered development process has introduced security or compliance risks because much of the model's behavior remains a "black box."

Enkrypt AI's research highlights the scale of the problem. Over the two months preceding the acquisition, the company scanned more than 268,000 AI tools across 25,000 MCP servers and identified over 143,000 vulnerabilities, affecting 73% of the servers it evaluated.

Unlike conventional cybersecurity threats, the companies argue that AI risks cannot simply be fixed through software patches. Every AI model, external tool and MCP server accessed by an autonomous agent potentially expands an organisation's attack surface.

"We red-teamed the world's leading frontier models and found exploitable vulnerabilities in every one. We shipped security across the entire agent stack. We earned the trust of enterprises most companies only dream of working with. We did all of this as a small team that simply refused to accept that security should be someone else's problem, solved after an agent had already shipped," Sahil Agarwal, Enkrypt AI CEO, said.

The startup has also evaluated models from major AI providers including OpenAI, Anthropic, Google Gemini, Mistral and DeepSeek, identifying exploitable attack categories across each platform.

With the acquisition, Enkrypt AI will be integrated into the Anaconda Platform alongside the company's existing AI orchestration capabilities and its recently acquired agentic coding platform Kilo Code.

"With Enkrypt AI, we're adding the agent security, governance, and compliance layer to the Anaconda Platform that makes sure all of it can be trusted. This means automated red-teaming across 300+ attack categories, real-time agent guardrails, and compliance controls mapped to frameworks like the NIST AI Risk Management Framework and the EU AI Act," David DeSanto, Anaconda CEO, added.

Anaconda said the combined platform will allow enterprises to discover, build, deploy, monitor and secure AI applications from development through production while providing the governance and audit capabilities needed for large-scale enterprise AI deployments.