Docker’s Open-Source Images Accelerate Secure AI Development: Anaconda CEO

Last week’s Docker announcement that it would make more than 1,000 of its Docker Hardened Images free and open-source software, combined with Anaconda’s AI Catalyst and other development technologies, will help overcome hurdles to building secure, scalable AI applications, Anaconda CEO David DeSanto tells CRN.

Docker’s recent move to make more than 1,000 of its container Docker Hardened Images free and open-source software marks a significant advance in efforts to accelerate AI software development and secure software development supply chains.

That’s according to David DeSanto, CEO of data science and AI development platform provider Anaconda, a leading Docker partner, who said in an interview with CRN that the newly open-sourced Docker container images, combined with Anaconda’s own offerings, are a breakthrough for AI development.

“One of the hardest things that developers face is knowing they’re using secure, trustworthy components. As a former software developer, I can relate to that. It’s only gotten even harder with the explosion of AI development and other things that have happened over the last decade. And so for developers, it’s important that they have a trusted ground to start from,” DeSanto (pictured) said.

“For us, we see [the Docker move] as a way to continue to reach out to developers and help them build secure AI-native applications, all from a trusted source, which would be Docker and Anaconda,” the CEO said.

While there has been a frenzy of AI software development in recent years, DeSanto said as much as 80 percent of AI projects never make into production. (An MIT Media Lab report in August said that only about 5 percent of generative AI pilot projects make it into production and achieve measurable value.)

Much of the problem is because developers lack secure, trustworthy components such as models and containers, DeSanto said. That slows down AI development projects, especially in the prototyping stage, as developers deal with vulnerabilities or the need to meet stringent data governance, security and sovereignty requirements.

Docker is a leading provider of cloud and AI-native development tools. It’s Docker Hardened Images (DHI), launched earlier this year, are pre-configured containers that have been fortified with built-in security measures that help secure the software development “supply chain”—the cycle of software applications from early-stage development through prototyping and operational production.

Last week, Docker announced that it is making its catalog of more than 1,000 Docker Hardened Images free and fully open source under the Apache 2.0 license. Docker said the move ensures that developers, software maintainers, hobbyists, development teams, governments and organizations “can use, share and build on DHI with clear rights and no hidden restrictions.”

(The images are available through the Docker Hub web site. Docker said businesses with unique requirements including customizations, regulated-industry compliance or accelerated patching can purchase Docker’s DHI Enterprise, with DHI Extended Lifecycle Support available for coverage after upstream support ends.)

Anaconda is a major player in the open-source software development space with its focus on the Python programming language. Anaconda’s Miniconda, the free installer for the company’s Conda package and environment manager for Python, is a standard part of Docker images.

Earlier this month, Anaconda applied its strategy of providing open-source development capabilities combined with data security, governance and compliance to the AI development world with the debut of its AI Catalyst suite of development tools for building, deploying and governing AI applications.

The core of AI Catalyst is a set of curated open-source AI models selected and vetted by Anaconda, along with risk profiles and other documentation.

The Anaconda-Docker Partnership

Anaconda’s partnership with Docker allows developers to combine Anaconda’s development environment management capabilities with Docker containers to accelerate software development and ensure AI and data science application portability.

“When you look at the Docker partnership, it’s about helping people start that journey. The hardest part of the AI journey is having an environment that you know will scale,” DeSanto said.

The CEO said one of the biggest challenges of moving applications from prototype to production is meeting an organization’s security trust model.

“There’s still this reluctance to be able to get AI securely into production. What this allows us to do for users, through our partnership [with Docker], is get [developers] to that trusted base that they can now develop on, that they know will be accepted in production because they’re using a Docker hardened image that meets their security team’s requirements. It will also speed up development, because it’ll be less time needed to configure the environment to begin the work” and later when troubleshooting the application, he said.

DeSanto said that with Anaconda and Docker, along with Anaconda’s integration with Nvidia’s GPUs, “You have a truly secure stack from your application down into the hardware.”

He noted that the Anaconda environment has 50 million users (and 2 million community contributors). “So if you look at it from that framework, we’re now enabling those 50 million users to be able to build more secure AI workloads faster as part of the partnership and potentially, by the way, reach past that 50 million to the additional [developers] who are also trying to do this and haven’t discovered Anaconda yet.”

The Anaconda-Docker alliance also benefits solution providers and ISVs who develop applications for their clients, helping them build AI software that can scale, meets security requirements, and avoids “the potential finger-pointing between infrastructure teams and developers,” DeSanto said.

“For us, it’s really about how we want to help people build what we’re calling trusted AI workloads at scale,” the CEO said, summarizing Anaconda’s intentions for 2026. “Another thing we’re focusing on is how to unlock more developer velocity. And the last part is just being secure by design. We’re looking at ways to provide more governance and controls so people can work within trusted guardrails to get their work done,” he said, pointing to the recent AI Catalyst launch.