Bringing the full power of agentic automation to the enterprises that couldn't go cloud first. A regional bank's loan-exception review queue, growing faster than the team can clear it. A federal agency's citizen-services backlog, where every case file contains data that legally cannot leave the country. A hospital system's prior-authorization workflow, backed up with documents that are subject to patient privacy law in every line. These are exactly the workflows that agentic AI was built to transform. And until now, they're also the workflows that agentic AI couldn't reach. The problem was never ambition. It was access. Most agentic AI platforms have been built on a single assumption: that the cloud is the operating model. Enterprise reality is more complicated—and for many of the world's most consequential workloads, it always has been. Public sector agencies operating under data sovereignty mandates. Banks bound by financial regulation. Hospitals safeguarding patient records. Defense and intelligence groups handling controlled information. Manufacturers protecting decades of intellectual property. The organizations and agencies with the most to gain from intelligent automation have had the least access to it.That changes with the April 2026 release of UiPath Automation Suite. The full UiPath agentic automation stack—the same AI agents, orchestration, and enterprise-grade capabilities are now available on-premises, on the Kubernetes infrastructure customers already operate on: AKS, EKS, and OpenShift. How it works Everything that matters to a regulated organization—the process logic that defines how work gets done, the agent configurations, the enterprise data being acted on, the audit trail that proves it happened correctly—runs inside the customer’s own Kubernetes environment and never leaves it. In self-hosted mode, the models themselves stay inside as well. The perimeter doesn’t move to accommodate the technology. The technology is built around the perimeter.A real demand, long underserved Agentic AI is no longer a future-tense conversation. Nearly two-thirds of enterprise leaders expect agentic AI to augment or replace many of their core business processes within the next two years. That ambition is widely shared. The ability to act on it has not been. More than 50% of AI leaders identify regulatory monitoring and infrastructure control as the most significant challenges in implementing sovereign AI. Data residency laws prevent certain information from leaving the country, the agency, or the perimeter. Chief information security officer (CISO) policies prohibit sensitive workloads from being routed to public large language model (LLM) endpoints. Compliance regimes demand the kind of auditability and control that cloud-only platforms haven't been fully able to guarantee. And many organizations have already invested in self-hosted, open-source models—with no enterprise-grade agentic platform to connect them to.The result is a paradox: the organizations running the highest-stakes processes, with the most complex compliance environments and the largest backlogs of work waiting to be automated, have been the least able to adopt the technology that could help them. This release closes that gap. What this opens up (and for whom) The UiPath Automation Suite April 2026 release brings agentic automation to the environments where it has been needed most. <strong>Public sector and government agencies</strong> can now apply intelligent agents to citizen-services workloads, benefits processing, and back-office operations—without changing the security posture that protects citizen data. Sovereignty mandates and in-country processing requirements are met by design, not by exception. <strong>Banking and financial services</strong> organizations can extend agentic capability across customer onboarding, exception handling, fraud review, and compliance workflows. Customer data, transaction records, and proprietary logic stay inside the perimeter where they belong. <strong>Healthcare and life sciences</strong> institutions can bring agents into claims processing, clinical operations, and research workflows. Patient privacy law is non-negotiable—and now it doesn't have to be a blocker. Defense, intelligence, and critical infrastructure environments gain access to agentic automation inside the boundaries that data classification and operational control require.<strong>Insurance</strong>, <strong>telecommunications</strong>, and <strong>manufacturing</strong> enterprises with strict in-country processing rules, IP-sensitive data, and CISO-enforced limits on public model usage now have a path forward they didn't have before. In every case: data that cannot leave the perimeter, infrastructure that must remain under direct control, and processes that have been waiting for an agentic option built around those realities rather than against them. What UiPath customers can now do On-premises AI has always forced a trade-off: capability or control. Full-featured agentic platforms required the cloud; on-premises meant settling for less. This release ends that trade-off. In practical terms, that means: Automate the work that couldn't leave the building. Sensitive processes (the ones tied to regulated data, controlled information, or sovereign requirements) are now in scope for agentic automation for the first time. Deploy the complete agentic stack, inside your own infrastructure. UiPath Maestro, Agent Builder in UiPath Studio, context grounding, ScreenPlay, Healing Agent, GenAI Activities—capabilities available in UiPath Automation Cloud™ are available here. Choosing on-premises means choosing control, not accepting a reduced feature set. Choose the model strategy that fits your governance posture. Run with cloud-hosted models from leading providers (OpenAI GPT, Anthropic Claude, Google Gemini), use UiPath-recommended self-hosted open-source models (GPT-OSS, GLM, Qwen), or bring your own. The choice is yours, and it stays yours.Govern agents the same way you govern the rest of your automation estate. Every agentic action runs within the audit, observability, and governance framework customers already use. Agents arrive as a first-class governed capability, not a parallel experiment to be contained later. The path that exists now The loan reviews. The citizen-services queues. The prior-authorization backlogs. The compliance workflows that have been quietly accumulating because no agentic option could reach them. These processes have not been waiting for ambition. They have been waiting for a path.That path now exists. And it’s expanding: the October 2026 LTS release adds Conversational Agents and UiPath IXP (Intelligent Xtraction Processing), continuing to close the gap between Automation Suite and Automation Cloud capabilities. <strong>See how Automation Suite deploys in your environment.</strong>For full feature availability details, refer to UiPath docs. Sources: Harvard Business Review, “Taming the Complexity of AI Data Readiness,” March 4, 2026. Deloitte, “AI Trends 2025: Adoption Barriers and Updated Predictions,” September 15, 2025.
![Image for [Highlight] Stephon Castle destroys Isaiah Hartenstein with the poster (with replay)](default-image.png)
.jpg)
