Red Hat launches ‘asago’ open-source community

Another day, another joint venture in the cybersecurity industry with a goal of bolstering enterprise AI security.

Earlier this month, Red Hat announced the formation of asago (short for AI Safety And Governance Orchestration), an open-source community that will help automate AI governance policies into “product-ready, safely deployed AI systems.”

Stuart Battersby, AI safety and model evaluation architect at Red Hat, told IT Brew that while the landscape of AI safety tools has matured, organizations still struggle with “slow processes” when translating AI policies into working software configurations.

“We’ve hit this inflection point where we have good AI safety tooling, but actually now what we need to do is make that AI safety tooling much more accessible to a wider business community that needs to be able to consume those things,” Battersby said. The purpose of asago is to help enterprises take AI policies to production “without needing to be experts in AI safety.”

“Our aim isn’t to replace existing AI safety tooling, but rather to act as an orchestration layer to join ecosystem components together,” the company wrote in an announcement of the project.

Collaborators of the asago community include tech giants such as Microsoft, Nvidia, and IBM Research, as well as academic institutions and government stakeholders.

Join the club. Asago is one of many efforts in the industry seeking to unite a group of tech giants in a bid to bolster AI security. Last month, IT Brew reported on the debut of the Open Secure AI Alliance, a coalition of companies dedicated to developing and circulating open-source AI security tools. Founding members included Amazon, Microsoft, and Okta. In 2024, the Coalition for Secure AI (CoSAI) formed to improve security around AI implementations, bringing together tech giants like OpenAI, Cisco, and Anthropic for the cause.

Grace Trinidad, IDC research director of AI security and trust, told IT Brew the rise of AI security industry groups and partnerships is due in part to the Hugging Face breach by a rogue OpenAI agent, which she called a “marquee event.” Both asago and the Open Secure AI Alliance were launched after the incident, which, she said, “just showed us how tricky these systems we’ve set up are actually becoming.”

Mike Toole, director of security and IT at security operations platform company Blumira, added that the influx of joint ventures stems from the realization that companies cannot defend against AI security threats in silos.

“One company’s not gonna be able to build a template for everyone,” Toole said. “Everyone’s making the effort to do that as an industry.”

Navigating the sea of joint ventures. Toole said more joint partnerships and communities like asago are likely to arise in the industry: “There’ll be a lot of these. It’s just some of them will succeed, some people will stop doing it. Just time will tell.”

Trinidad said IT pros should be watching the space and assessing the weaknesses and strengths of each framework.

“The worst thing that an IT pro can do is to assume that any one of them is better than the other,” Trinidad said. “I don’t think that we’re ready yet to say that this approach is superior or that this definitely meets the needs of everyone.”