Artificial Intelligence (AI) is being talked about everywhere, the software solution providers are shouting about AI in processes everything from invoice automation to fraud detection. But as AI becomes a must have for companies adopting a digital transformation there is a growing number of solution providers are guilty of AI-washing. Much like “greenwashing” in sustainability, AI-washing refers to companies overstating or misrepresenting the role of AI in their products. It’s a tactic used to attract attention, justify pricing, or appear innovative without delivering real intelligence.
For finance teams managing Accounts Payable (AP) and Purchase to Pay (P2P) processes, this matters more than ever. Companies that are undergoing digital transformations are looking to invest in tools to streamline operations, reduce errors, and improve compliance. It is essential that the functionality of these solutions enables these objectives to be met. Buying into a solution that offers AI to help with these processes and then doesn’t deliver leaves the organisation (and team) with a hole in their processes and a software solution that doesn’t deliver.
What Does AI-Washing Look Like in Finance Solutions?
AI-washing in AP and P2P platforms is often presented as:
- Generic claims like “AI-powered invoice processing” without explaining how AI is used.
- Rebranding rule-based automation as “machine learning” or “intelligent workflows.”
- Overpromising capabilities like fraud detection or predictive analytics that don’t work in practice.
- Buzzword-heavy marketing with little technical substance or transparency.
In AP and P2P, accuracy, speed, and compliance are a critical requirement to enable streamlined processes for digital transformation. Having the ability to identify solutions that are offering real AI value is essential in the evaluation and selection of a software provider.
Here is a brief guide to some questions that will help identify the AI that will benefit your organisation’s P2P and AP processes:
- What specific AI technologies are used? Is it machine learning, natural language processing, or something else? Ask how the AI improves processes such as invoice matching, fraud detection, or approval routing?
- Is the AI embedded in workflows? How does it enhance invoice capture, duplicate detection, or spend analysis?
- Ask for real-time examples? Case studies or metrics showing how AI has improved processing times, accuracy, or compliance. For instance, invoice capture on first pass and then improvement metrics on second and third pass etc
- Is the AI explainable and auditable? Can you trace how decisions were made and justify them during audits? Do they have in-house AI expertise? A credible provider will have an in-house development team that developed the code for the AI functionality, works with the Account management teams to develop a road map linked to customer feedback and new technology for continuous AI improvement
What Real AI Looks Like in AP/P2P
Genuine AI in finance solutions should deliver measurable value. For example:
- Smart invoice capture that learns from historical data to improve accuracy.
- Automated GL coding based on context and past transactions.
- Duplicate and fraud detection using pattern recognition and anomaly detection.
- Predictive analytics for cash flow forecasting or supplier risk.
These features require robust models, quality data, and ongoing development and updates.
Don’t Forget the Humans in the Loop
Even the most advanced AI in Accounts Payable and Purchase to Pay systems needs human oversight. AI can accelerate invoice matching, flag anomalies, and suggest GL codes but doesn’t understand business context, supplier relationships, or exceptions in the way the team does.
Having humans in the loop ensures that AI decisions are reviewed, validated, and refined. After all, machine learning as an AI tool is only as good as it’s teacher – the human! It allows finance teams to catch errors, override incorrect suggestions, and provide feedback that helps the system learn and improve. This collaboration between human expertise and machine intelligence is what drives real efficiency and trust.
AI has the potential to accelerate digital transformation in AP and P2P processes but only if it’s real, relevant, and responsibly deployed. Challenge providers, ask the right questions, and ensure the tools that are promoted are fit for purpose.
