OpenText Commits €105 Million to Ireland, Doubling Down on Sovereign Cloud and Agentic AI

OpenText will invest €105 million ($120.4 million) in Cork and Galway over the next three years, creating 400 jobs in what it calls the single largest investment in Ireland by a Canadian-headquartered technology company. For SAP teams, the news is less about geography than architecture. It highlights whether agents can act on trusted SAP and non-SAP information within the right residency, governance, and integration model.

OpenText’s expansion, supported by the Irish government through IDA Ireland, was unveiled alongside Taoiseach Micheál Martin TD and IDA Ireland CEO Michael Lohan. The company framed it as the largest single investment in Ireland by a Canada-headquartered technology company.

The money funds operations and R&D across three areas that OpenText calls foundational to trusted enterprise AI: agentic AI, sovereign cloud, and cybersecurity. According to OpenText, Irish-based developers and researchers in the two cities will “design, deploy, secure and operate these AI & cloud capabilities for EMEA markets,” with agentic AI work spanning multi-agent collaboration, system boundary enforcement, and knowledge sharing across sovereign zones.

Shannon Bell, EVP, Chief Digital Officer and CIO at OpenText, tied the investment to demand. “Organizations across Europe are looking for trusted partners that can help them deploy AI securely, govern it responsibly, and operate with confidence across increasingly complex digital environments,” she said. That framing matters because data residency and AI enablement have often been evaluated separately, one through compliance and infrastructure, the other through business use cases and process automation.

Sovereign Capacity Meets the Information Layer

OpenText’s announcement treats agentic AI, sovereign cloud, and cybersecurity as related architectural concerns rather than separate purchases. For a European SAP customer weighing AI agents across SAP and non-SAP systems, several questions converge. That includes where the data resides, how it is classified, which systems the agent can access, and whether the resulting actions can be monitored or audited.

In that context, secure information management for AI is less a product label than an operating layer. It can influence whether AI is usable in regulated environments, but it does not remove the customer work required to make information usable.

The investment also fits a broader cloud shift. SAPinsider Cloud Research shows that cloud-native and hybrid SAP architectures are driven by faster innovation and deployment at 45%. Composable ERP/cloud requirements, technical debt reduction, and cost/data center optimization each account for 32%. Key technologies cited include cloud data lakes and data services at 45%, integration platforms at 41%, data-integration/data-mesh capabilities at 41%, and cloud AI/ML tools at 38%.

Demand Is Still Ahead of Readiness

OpenText is investing in a market where interest in agents is outpacing implementation. SAPinsider AI Research shows that 14% of organizations are currently implementing AI agents, 23% plan to do so within 12 to 24 months, 30% are evaluating or are aware of the need, and 20% have no plans.

The maturity gap is significant. Agentic AI adoption is highest among Data Leaders at 37%, compared with 18% for Data Adopters and 13% for Data Beginners. Data Leaders also report higher use of embedded analytics, natural-language querying, synthetic data/digital twins, and predictive AI/ML.

That gap explains why the sovereign cloud portion of the announcement may matter sooner for some organizations than the agentic AI portion. SAPinsider Data and Analytics Research finds that 79% of organizations cite data quality, security, and resiliency as requirements, while 73% cite regulatory compliance. Those are current requirements. AI agents may shape longer-term planning, while secure, regionally governed information capacity addresses problems customers already face.

Governance Remains the Customer’s Work

Sovereign cloud capacity is one element of the architecture, not the architecture itself. AI agents are not safer because data is located in a particular geographic region. Residency addresses jurisdiction. It does not determine whether a process agent has appropriate entitlements, whether content has been classified, or whether the lineage of a master data record can be reconstructed. This is where residency can become a jurisdictional comfort blanket rather than a model of control.

SAPinsider research suggests many organizations are still building those foundations. Only 38% report implementing new governance, quality, and compliance policies, and the same 38% are building an enterprise-wide data strategy. Those figures should matter before organizations delegate meaningful decisions to autonomous agents.

What This Means for SAPinsiders

Treat residency, classification, and agent entitlement as one design. OpenText’s Ireland build-out makes sovereign cloud and AI-ready information capacity easier to evaluate, but it does not change the underlying data work. A sovereign zone indicates where data lives; it says nothing about whether a process agent has the appropriate entitlements, whether content is classified, or whether a master data record’s lineage can be reconstructed. SAPinsiders should collapse data residency, classification, and agent access design into a single architecture workstream across SAP and non-SAP estates, and validate it before any agent touches production data.

Test agentic AI business cases against real governance maturity first. The market signal is loud, but readiness is not. An agentic business case built on assumed automation value will stall against the data and control foundations that CIOs have, not the ones they wish they had. Organizations must, therefore, stage agentic investment behind a governance-maturity checkpoint and prioritize the sovereign cloud and data-quality work that addresses problems the organization already faces this year.

Position sovereign-cloud integration as a data-engineering engagement. European client requirements, reinforced by regulatory compliance, will increase demand for sovereign cloud patterns spanning SAP and non-SAP systems. The binding constraint for these clients is not infrastructure capacity but trusted, integrated, auditable data. Lead with data engineering, classification, and lineage as the central dependency, and frame OpenText-style capacity as enabling, not sufficient, in client roadmaps.