OpenText and Cohere target regulated AI as $4.25bn ARR base seeks faster growth

Open Text Corporation (NASDAQ: OTEX; TSX: OTEX) and Cohere have formed a strategic partnership aimed at governments and regulated industries that want agentic artificial intelligence without sending sensitive information into conventional public AI environments. OpenText will provide the enterprise data and context layer across unstructured, operational and transactional information, while Cohere contributes its privately deployable North platform and enterprise AI models for reasoning, orchestration and automation. The companies did not disclose a contract value, customer commitment or revenue-sharing arrangement, so the partnership is not itself a material booked-sales event. Its importance lies in OpenText attempting to turn decades of enterprise information management into infrastructure for AI agents, while Cohere gains access to customers whose main obstacle to AI adoption is often governance and data control rather than model capability.

OpenText enters the relationship with fiscal 2026 annual recurring revenue of $4.246 billion and full-year revenue of $5.246 billion, but total revenue increased only 1.5%. Cloud revenue grew faster at 5.5% to $1.959 billion, and enterprise cloud bookings increased 22.5% to $947 million, suggesting the company has improving forward momentum in precisely the areas that could benefit from secure AI deployments.

Why could enterprise data matter more than the AI model for regulated customers?

Large language models are increasingly accessible from several suppliers. The harder problem for enterprises is giving those models reliable access to internal information while ensuring the agent sees only what the user is authorized to see and acts within established business rules.

OpenText already manages documents, communications, workflows and operational information across large organizations. That creates a potential data layer through which an AI agent can retrieve context rather than operating primarily on public information or manually copied prompts.

Cohere’s North platform supplies the intelligence and orchestration layer while supporting private deployment. This matters in government, finance and other regulated industries where customers may not want sensitive data flowing into a public AI service even when the underlying model provider offers strong contractual protections.

The combination therefore targets a bottleneck different from raw model performance. An enterprise agent that reasons brilliantly but cannot securely access the documents, policies and transactions needed to perform work has limited commercial value.

Why is sovereign AI becoming a meaningful enterprise software category?

Sovereign AI generally refers to keeping control over models, data, infrastructure and governance within jurisdictions or organizational boundaries that customers consider acceptable. The concept has gained importance as governments worry about depending on foreign cloud and model providers for sensitive public-sector workloads.

Regulated companies face similar constraints even when national sovereignty is not involved. Banks, insurers, healthcare companies and utilities may need to control where data resides, how it is processed and which employees or automated systems can access it.

Cohere has positioned itself aggressively around this market by offering models and North in privately deployable configurations rather than insisting that every workload run through a centralized proprietary cloud. OpenText can strengthen that proposition because customers also need the information-governance layer surrounding the model.

The challenge is that sovereign deployment can be more expensive and technically complex than consuming a public AI API. Customers must believe privacy, control and regulatory compliance create enough value to justify that additional operational burden.

Can AI reaccelerate OpenText after fiscal 2026 revenue grew only 1.5%?

OpenText’s fiscal 2026 results show a mature software company with strong cash generation but limited overall growth. Full-year revenue increased 1.5% to $5.246 billion, ARR increased 1.3% to $4.246 billion and cloud revenue rose 5.5% to $1.959 billion. Adjusted EBITDA reached $1.903 billion at a 36.3% margin, while free cash flow increased 17.5% to $808 million.

The partnership with Cohere is strategically appealing because it can add new AI workloads without OpenText having to develop a frontier model independently. OpenText can focus on data, governance, security and enterprise integration while Cohere supplies the model and agent platform.

That structure can improve capital efficiency relative to competing directly in frontier-model training, where infrastructure costs can reach billions of dollars. It also protects OpenText from betting the entire product roadmap on one internally developed model architecture.

The difficult part is monetization. Customers may expect AI functionality to become part of existing software subscriptions rather than paying large incremental amounts, so OpenText needs clear pricing around higher-value automation or new cloud consumption.

Why do OpenText’s enterprise cloud bookings matter more than the partnership announcement alone?

Enterprise cloud bookings reached $947 million in fiscal 2026, rising 22.5%, while fourth-quarter bookings increased 24.1% to $295 million. Those numbers measure new committed incremental cloud contracts and therefore provide a stronger signal of future business momentum than the 1.5% full-year revenue growth rate alone.

If Cohere integration helps OpenText win more of those contracts, AI can contribute to an already improving sales channel rather than requiring management to construct an entirely new business.

OpenText also carries substantial recurring revenue, meaning new AI capabilities can potentially be cross-sold into existing enterprise relationships. That is easier than acquiring every customer from scratch, particularly in regulated industries where vendor qualification can take months.

The risk is that installed customers increasingly compare OpenText with newer AI-native data platforms. Legacy relationships become an advantage only if the company can modernize quickly enough that customers do not move information and workflows elsewhere when adopting AI.

What does OTEX’s stock performance say about investor expectations?

OpenText’s Nasdaq shares closed September 18 at $22.58, down 4% for the session. The stock was approximately 2.1% below its September 11 close of $23.06 and about 6.6% below the August 18 close of $24.17, while its latest 52-week range stood at $19.78 to $39.90.

Those numbers show that investors remain cautious despite strong cash generation and accelerating enterprise cloud bookings. At roughly $5.5 billion of market capitalization in recent trading, OpenText is valued very differently from faster-growing enterprise AI software companies.

That discount creates an opportunity if AI can reaccelerate organic growth without damaging margins. It also reflects the burden of proof facing management after years in which acquisitions, restructuring and portfolio changes made the growth profile difficult to interpret.

The Cohere agreement matters because it gives OpenText a credible model partner without requiring frontier-model spending. It becomes financially meaningful only when sovereign AI starts showing up in bookings and recurring revenue.