Enkrypt AI spent the two months before this deal doing something that sounds almost boring until you see the number attached to it. It scanned more than 268,000 tools across 25,000 MCP servers, the connectors that let AI agents plug into outside services and actually go do things. It found over 143,000 vulnerabilities. That touched 73 percent of the servers it looked at.
Anaconda bought the company this week. If you don’t touch data science tooling day to day, Anaconda is the platform a huge share of Python and machine learning teams already build on, so this isn’t some outside security vendor trying to force its way in. It’s the plumbing provider deciding to add a lock to the pipes it already owns.
What Enkrypt brings is red-teaming before an agent ships, guardrails while it’s running, and the compliance paperwork that maps to rules like the EU AI Act, the kind of work nobody wants to do by hand. Folded into Anaconda’s platform, a team building an agent can get flagged for a bad dependency or a leaky permission before it reaches a customer, instead of finding out after something breaks.
This is Anaconda’s second security-adjacent purchase in as many months, after picking up the coding agent Kilo Code in July. Put the two together and the pattern is easy to read. Companies are racing to ship agents fast, and the tooling underneath them is racing just as hard to keep up with what those agents are allowed to touch.
I like seeing a deal like this get made quietly instead of after something goes wrong. Most of the scary AI security stories start with a gap nobody checked ahead of time. This is somebody checking first, and building the checking into the tool people were already using.
——
Follow: @Ali Demi
Book your free AI clarity call, NOW!
https://buff.ly/TpWy277
——
Sources:
https://www.anaconda.com/blog/anaconda-acquires-enkrypt-ai
https://www.hpcwire.com/aiwire/2026/08/04/anaconda-acquires-enkrypt-ai-to-secure-the-trillion-token-enterprise/
Repost this. Thanks.

