Basware Establishes Governed Autonomy Framework for Finance

New operating model gives AI authority while maintaining human control

CHARLOTTE, N.C., June 30, 2026 /PRNewswire/ -- Boards want AI scale. Auditors want accountability. Finance teams need both. Yet only 36% of CFOs feel confident in their ability to deliver real enterprise impact from AI initiatives, according to Gartner research.  

Today, Basware, the global leader in Invoice Lifecycle Management (ILM), announced the Governed Autonomy Framework for Finance: a new operating model that lets organizations expand AI authority within customer-defined controls, policies, and audit trails, so that autonomy, accountability, and financial integrity advance together.  

"Every CFO I talk to has the same problem," said Jason Kurtz, CEO, Basware. "Their board wants them to use AI. Their auditors want a trail of accountability for every decision that touches the books. Their teams want to know exactly what AI is allowed to do and where they still have the final word. Until now, no architecture in our category answered all three at once. That is what Governed Autonomy is built to do- make AI in finance defensible, not just deployable."  

Inside finance, AI is moving from assistance to execution, reviewing invoices, resolving exceptions, and increasingly, acting on behalf of the team. In a survey of global finance leaders, 92% of CFOs said they are already integrating AI into financial decision-making, either across some processes or virtually all of them.  

But existing approaches force a trade-off: highly automated systems create compliance risk when decisions cannot be explained or traced, while highly controlled systems give up the benefits of automation. Only 12% of business leaders say AI has delivered both cost and revenue benefits, while 56% report no significant financial benefit yet, according to PwC research. Finance leaders need AI authority to grow without surrendering governance, compliance, financial control, or the financial integrity of the business.  

"Finance doesn't get the luxury of using technology just because it's impressive. The question isn't whether AI can do the work - it clearly can. The question is whether you can trust the outputs, govern the process, and stand behind the results when they're challenged. We weren't looking for AI that bypasses controls; we were looking for AI that operates within them. That's what turns AI from an interesting capability into something finance can actually use," said Tehreem Bokhari, VP of Finance Transformation at NOV.  

A Structured Path from AI Assistance to AI Execution  

Basware's Governed Autonomy Framework defines three levels of AI authority, each set by the customer rather than the vendor. The customer decides where automation begins and where human oversight remains, on a process-by-process basis.  

1. AI as Adviser - The system recommends, the human decides.  
2. AI as Collaborator - The system acts within agreed thresholds and policies, the human reviews outcomes and exceptions.  
3. AI as Operator - The system runs autonomously within customer-defined controls and confidence levels, with every action logged, traceable and auditable.  

As confidence grows and outcomes are validated, organizations expand AI authority progressively, without reimplementation, and without breaking audit continuity.  

"Autonomy without governance is reckless. Governance without autonomy is just expensive software," Kurtz said. "The question facing every finance leader is no longer whether AI can do the work, but how much authority they are willing to give it, and whether they can defend every decision afterward. Governed Autonomy lets organizations scale AI adoption without losing visibility into how decisions are made or who remains accountable for the outcome."  

Governance Built into Execution, not Bolted on  

Basware's Governed Autonomy Framework for Finance rests on a simple principle: AI operates only within the authority finance teams choose to grant it. Every recommendation, action, escalation and exception is governed by customer-defined policies, approval thresholds, compliance requirements and audit controls.  

Consider a real three-way match exception. A $480,000 invoice arrives from a long-standing supplier. The purchase-order quantity matches. The receipt is confirmed. But the unit price is 4.3% above the contracted rate, which is inside a tolerance range some companies allow and others do not. In a conventional system, that exception drops into a human queue, gets approved or rejected, and the rationale lives in an email thread nobody can find later.  

Basware routes the exception, flags the variance, applies the customer's pre-configured tolerance policy, recognizes that the supplier has a clean 36-month payment history with no prior pricing disputes, and then, depending on the threshold the finance team has set, either resolves the exception automatically with a full decision trace or escalates it with a structured recommendation. Every step is logged. Every threshold is customer-defined. Every recommendation can be reviewed, adjusted, or overridden.  

Built on Four Decades of Invoice Intelligence  

Governed AI without deep data is just a well-intentioned policy document. Basware has processed invoices, and the compliance obligations that surround them, for more than four decades, for over 6,500 enterprise customers. The platform is trained on more than 2.5 billion invoices and supports more than one billion AI actions annually. It has seen most of the patterns that cause exceptions long before any single customer encounters them, and those patterns inform how finance teams set and refine their thresholds.  

Finance teams use that accumulated intelligence to configure AI behavior on their own terms, defining when AI advises, when it collaborates, and when it operates without review, then expand those thresholds as confidence builds, without re-implementation or complex reconfiguration, and without ever losing auditability.  

"Everyone is working to teach AI what it can do. The harder problem is teaching it what it must never touch," said Donna Wilczek, Chief Product and Technology Officer at Basware. "The market is focused on autonomous AI only. We are focused on accountable autonomy. A mandate in France is fixed in law, and it is not the same as Poland's or Germany's. An AI that doesn't understand those boundaries is not autonomous. It's unsupervised. Compliance is not a feature we add after the decision. It is the foundation that defines what the AI is allowed to do in the first place."  

Compliance as the Foundation, not a Layer  

The timing is not incidental. France's e-invoicing reforms, Poland's KSeF framework, Germany's B2B e-invoicing requirement, and the European Union's VAT in the Digital Age (ViDA) initiative are placing new technical and legal obligations on global enterprises, often across multiple jurisdictions at once. Basware's compliance coverage spans more than 60 country mandates and is maintained continuously as mandates evolve.  

Critically, that compliance layer is not applied on top of the AI system. It is the foundation on which Basware's AI operates. Governance applies the same way to every workflow rather than being switched on selectively for AI features, and the limits on autonomy are the ones each customer sets through its own confidence thresholds, risk profiles, and process design. Accountability stays with people, while the system supplies the decision trace, the explanation, and the system-level evidence behind every outcome.  

Defining the Next Era of Finance  

Finance automation started by digitizing work. The shift underway now is to delegating it, and the real test is not how much autonomy a system has, but how much of it finance teams can account for afterward.  

"The organizations that win the next decade will not be the ones that deploy the most AI, but the ones that can govern it, trust it, and use it to strengthen financial integrity across every transaction and every decision," Wilczek said. 

About Basware  

Basware is how the world's best finance teams gain complete control of every invoice, every time. Our Intelligent Invoice Lifecycle Management Platform ensures end-to-end efficiency, compliance and control for all invoice transactions. Powered by the world's most sophisticated invoice-centric AI, trained on over 2.5 billion invoices, Basware's Intelligent Automation drives real ROI by transforming finance operations. We serve 6,500+ customers globally and are trusted by industry leaders including DHL, Heineken and Sony. Fueled by more than 40 years of specialized expertise with $10+ trillion in total spend handled, we are pioneering the next era of finance. With Basware, now it all just happens. 

Media Contact 

Abs Hassanali  
Global Head of PR & Communications  
[email protected]  

Photo - https://mma.prnewswire.com/media/3001824/Artboard.jpg
Logo - https://mma.prnewswire.com/media/2398888/6017253/Basware_logo.jpg

SOURCE Basware