Red Hat wants to be the control tower of AI agents and reinforces RHEL and OpenShift
At KubeCon, Red Hat announced improvements in RHEL and OpenShift to manage and protect AI agents created on different platforms such as AWS, Azure, Google Cloud, Salesforce...

Key takeaways
- Na KubeCon, a Red Hat anunciou melhorias no RHEL e no OpenShift para gerenciar agentes de IA.
- A Red Hat quer ser a 'torre de controle' de agentes de IA criados em AWS, Azure, Google Cloud e Salesforce.
- O anúncio ainda não resolve todos os desafios de controle e proteção de agentes de IA.
- A estratégia da Red Hat se diferencia das abordagens de AWS, Azure e Google Cloud para agentes de IA.
At KubeCon, Red Hat presented improvements in Red Hat Enterprise Linux (RHEL) and OpenShift aimed at managing and protecting artificial intelligence agents in corporate environments. The company wants to occupy the role of "control tower" of these agents — a layer capable of observing, authorising and limiting what autonomous systems do, even when they were built outside the Red Hat ecosystem. The movement responds to concrete pain: AI agents are born today in AWS, Azure, Google Cloud, Salesforce and OpenAI, and rarely talk to each other under a single security policy.
Why did controlling AI agents become a priority?
Until recently, the corporate challenge was to put models into production. With agents able to execute actions — call APIs, open calls, move data, fire transactions — the risk changes in nature: it is no longer "the template has missed text" and becomes "the agent has done something wrong in a real system". When a company operates dozens of agents on different platforms, each supplier brings their own permissions and logs panel. The result is a puzzle in which no one sees the whole, and incidents become difficult to trace. According to Network World coverage, this is exactly the fragmentation Red Hat is trying to solve.
What changes in RHEL and OpenShift for AI agents?
The logic of the ad is to use the basis that already supports much of the corporate infrastructure as a control point. RHEL, an operating system in which many workloads run, and OpenShift, a Kubernet-based container orchestration platform, are now incorporating management and security resources designed for AI agents. The competitive advantage Red Hat tries to sell is neutrality: instead of demanding that the customer standardize everything in a single cloud provider or model provider, the company positions itself as the common link between heterogeneous agents. It is the same dispute in which integrators and hyperscalers are moving — SoftwareOne, for example, took your AI agent engineering to AWS Marketplace with the AI-DLC method(/noticias/softwareone-engenharia-agentes-ia-aws-marketplace).
How does Red Hat differ from AWS, Azure and Google Cloud in AI agent control?
Each hyperscaler offers robust governance — but by definition within the cloud itself. Red Hat's bet is the opposite: a control plan that runs through environments. This dialogues with the company's open source heritage and the reality of companies that keep workloads on-premise, in private clouds and in more than one public provider at the same time. For the platform team, the promise is to reduce the number of consoles that need to be monitored and focus access policies on one point.
What does control of AI agents mean for those operating AI in Brazil?
Brazilian companies that already deal with agents in production face the same visibility problem described at KubeCon — aggravated by LGPD, which requires knowing who accessed which personal data and for what purpose. An agent without a consolidated audit trail is a regulatory liability. The case of health is illustrative: [7 in 10 organizations in the sector have already lost control of patient data because of shadow IT with AI. Centralised control tools, if they work as promised, attack precisely that point.
What the announcement of AI agents still does not solve
There are clear limits. The ad does not set a market standard for agent identity, and Red Hat did not disclose deadlines, prices or performance metrics for new capabilities. It is also unclear how the control will work for agents running entirely within closed third party platforms, without component running on RHEL or OpenShift. The dispute for being the corporate AI traffic hub is just beginning — and the winner will be the one to convince customers that he can see all agents, including those who are not his. The full coverage is in Network World.
Frequently Asked Questions about AI and Red Hat Agents
What did Red Hat announce at KubeCon?
Improvements in RHEL and OpenShift aimed at managing and security of AI agents in corporate environments, with central control positioning for agents built on multiple platforms.
Why do companies need control over AI agents?
Because agents perform real actions on corporate systems. Without unified visibility and policies, it is difficult to audit accesses, contain errors and meet regulatory requirements such as LGPD.
Will Red Hat replace AWS, Azure and Google Cloud tools?
Nope. The proposal is to act as a common layer between different environments, not as a substitute for the native tools of each cloud provider.
Source: www.networkworld.com

Written by
Lucas MontarroiosSou Lucas Montarroios e dediquei os últimos 15 anos à linha de frente de operações de telecom e data centers. Minha carreira sempre foi pautada por um foco implacável: transformar tecnologia e cenários críticos em oportunidades reais de negócio. No novidades.ia.br, trago essa visão executiva para o universo da IA. Especialista em produtos, mercado e ferramentas práticas de IA. Minha missão aqui é filtrar o ruído do mercado, analisando benchmarks, estratégias de grandes empresas e ferramentas práticas.


