Red Hat: ‘The power of AI is open’

  • The IBM-owned provider champions open-source as ideal for AI innovation, fostering rapid, secure, and compliant AI development and deployment

  • The crimson cap company underscores its strong connections with open-source communities and partnerships to deliver comprehensive AI solutions that can operate across multiple cloud and on-premise environments

  • The ruby bonnet vendor is pushing open source amid a broader debate on open-source versus proprietary AI software security

Red Hat issued a full-throated endorsement of open source as optimally suited to delivering on AI’s requirements for rapid development, widespread deployment, multi-platform, multi-vendor integration, security and regulatory compliance.

“At Red Hat, we believe open source communities are where innovation begins, and this is especially true when it comes to AI,” Ashesh Badani, Red Hat senior vice president and chief product officer, says in a blog post titled “The Power of AI Is Open,” and scheduled to be posted Monday, during the KubeCon + CloudNativeCon North America conference in Chicago.

And of course, according to Badani, Red Hat has the chops needed to deliver on open source for AI: “Our deep-rooted relationships with open source communities such as Kubeflow, KServe, CodeFlare and Ray, open source models from companies like Hugging Face and our upstream first approach allows us to foster open, transparent contributions, leading to downstream innovation with our partners.”

Additionally, Red Hat’s collaborative relationships with hardware and software partners help customers build a more complete AI solution, he says. Teams can use existing tools and processes to enable quick work to get AI-enabled applications into production while meeting security, regulatory compliance and governance standards, as organizations do with other business applications.

The company accelerates time to value for AI, bringing together data scientists and developers, Badani says.

Red Hat goes 'anywhere'

And Red Hat helps companies deploy AI-enabled applications anywhere, “across any cloud—public, private, on-premise or at the edge—at scale and with greater security capabilities,” Badani says.

Red Hat can support organizations concerned with data privacy, gravity and sovereignty by moving AI models closer to data sources, helping build and use AI models in alignment with regulatory requirements restricting data movement, Badani says.

Of course, Red Hat touted its own technology as great for implementing scalable, secure and regulatory-compliant AI: OpenShift AI and OpenShift help development, security and operations teams automate and streamline the development and deployment process, train in the cloud and bring the model to the private data center for greater GPU power. Ansible and IBM watsonx automate operations and close the IT skills gap, Badani says.

The company touts customer success stories: HCA Healthcare uses predictive analytics to more accurately and rapidly detect sepsis, and NTT East is developing a video AI service for intelligent business insights, reducing service operating costs by 50-60%, with improved management and version control.

Red Hat’s openness message is targeted to win more mindshare across the developer and data scientist community in driving AI innovation, Futurum Group analyst Ron Westfall said. The company faces stiff competition from the major cloud providers, including Amazon Web Services, Microsoft Azure, and Google.

Better together

Major AI platforms are “fundamentally open source,” though they must interoperate with private AI models developed by organizations for privacy and proprietary data access protection, Westfall said.

Red Hat OpenShift AI is a leading open-source platform for building, training, testing and serving models, the analyst said. The vendor is building on OpenShift’s global presence and vast channel influence supporting developers building applications, especially across containers and Kubernetes environments. And Red Hat benefits immensely from OpenShift alliances, including IBM Watson Studio, Anaconda, Intel OpenVINO and NVIDIA AI enterprise, Starburst, and 30 more AI and ML certified partners, Westfall said.

AI revives an old controversy about the relative security of open source vs. proprietary software. Advocates of proprietary software claim that private companies are best positioned to secure their products. Keeping source code secret makes attackers less likely to discover security flaws. Meanwhile, open source advocates adopt the principle that “given enough eyeballs, all bugs are shallow” — large communities of open source developers will find and fix bugs (and security flaws) before attackers can find them.

AI heightens the stakes, with the risk that evildoers might harness AI for terrorism, oppression, fraud, threats to critical infrastructure—and that AI itself might evolve, Terminator-like, to jeopardize humanity.

Last week, US President Joe Biden issued an executive order directing agencies to develop safety guidelines.

Red Hat is in the open-source business, so it’s unsurprising that it falls on the side of open source. However, this week’s statement focuses on development tools and infrastructure and does not address the AI itself.

For that, look to Red Hat parent IBM, which is pushing hard to lead enterprise AI, launching watsonx, a platform for building AI for business.

Want to discuss AI workloads, automation and data center physical infrastructure challenges with us? Meet us in Sonoma, Calif., from Dec. 6-7 for our Cloud Executive Summit. You won't be sorry.