Unit4 offers mid-market AI trials without commitment

Unit4 has launched an initiative that allows new and existing ERP customers to try its AI tools without a subscription commitment. The offer is aimed at the mid-market.

Eligible customers that register before the end of December 2026 will get access to AI features across the Unit4 ERPx platform. They can use the tools until the end of August 2027, then either take out a subscription or stop using them.

Under the scheme, customers can use Unit4's Advanced Virtual Agent, Ava, which responds to natural language requests through tools including Microsoft Teams. The software is designed to work across finance, HR and people, procurement, and project management processes.

The offer applies to both new and existing ERP customers and includes a fair use cap intended to support regular use without extra charges. Unit4 is targeting organisations in service-focused sectors that already use, or are considering, its ERP system.

The launch reflects a wider push by software suppliers to reduce barriers to AI adoption among medium-sized businesses, many of which have been slower to commit spending than larger companies. Vendors have increasingly used trial access and limited-risk pricing to encourage testing before longer-term contracts are signed.

Simon Paris, Chief Executive Officer at Unit4, said the company wants mid-sized organisations to test AI without being tied to a purchase decision.

"The mid-market has always been an engine of growth for economies around the world, and we want to help build a new generation of AI-enabled mid-market champions," said Simon Paris, Chief Executive Officer, Unit4.

"The 'AI for Your World' initiative underlines our commitment to helping our customers experiment with AI functionality without concerns about contractual obligations. We believe this commitment-free approach will trigger faster innovation and productivity, giving teams time back for the work that matters most," Paris said.

AI approach

Unit4 set out four principles behind its AI strategy: ERPx has been built with sector-specific workflows; governance controls are embedded in the architecture; the system improves through repeated use within each tenant and jurisdiction; and the design is intended to avoid reliance on a single large language model.

Actions taken through the system remain subject to customer rules, thresholds, and delegation structures, with each action attributable to a named human. Unit4 added that its data protection measures include tenant isolation, data residency, and a policy of not using customer data to train public large language models.

Claus Jepsen, Chief Technology Officer at Unit4, said AI will be effective in this part of the market only if it is built into core business systems rather than added separately.

"AI will only be effective for mid-market organisations if it is embedded in an intelligent core that is fluent in the language of our customers' industries," said Claus Jepsen, Chief Technology Officer, Unit4.

"They have a head start using our AI with ERPx because it has been designed with their needs in mind. It becomes smarter the more involved it gets in work, ensuring institutional memory lives in the system and accelerates productivity. It is auditable, data is protected, and our customers can be confident of compliance with regulatory frameworks. It is also model-agnostic, avoiding lock-in to one LLM approach," Jepsen said.

For customers, the practical question is whether a no-commitment trial will be enough to move AI projects beyond pilots and into routine operational use. In many mid-sized organisations, decisions on automation and AI adoption remain tied to concerns about governance, cost control, integration with existing processes, and whether staff can use the tools without extensive retraining.

By placing the offer inside its ERP environment, Unit4 is trying to make AI part of day-to-day administrative work rather than a separate experimental layer. Ava's role inside collaboration tools such as Microsoft Teams points to a model in which employees ask for information or trigger tasks in natural language instead of moving between multiple business applications.

That could be relevant in sectors where finance, project oversight, procurement, and workforce management are closely linked, and where administrative work often spans several systems and teams. Unit4's pitch is that AI should sit within those workflows rather than outside them, with controls that allow organisations to monitor how tasks are handled and by whom.

Customers that sign up before the registration deadline can stop using the tools at any point before the programme ends on 31 August 2027.