AI VAT product strategies are beginning to sound very similar.
That fails to understand the opportunity
Every tax technology vendor now seems to have an AI strategy. Most demonstrations look remarkably similar:
- An AI assistant answers VAT questions.
- A chatbot explains a transaction.
- An agent prepares a VAT return.
Interesting, perhaps. Transformational? Not really.
The uncomfortable truth is that preparing a VAT return has never been the hard part. The hard part is finding the right data in the first place.
It’s the data gap, stupid
Most multinational businesses operate multiple ERP systems, inconsistent tax code structures and years of legacy configuration. By the time the VAT team receives the data, the problems have already been created. Missing fields, incorrect mappings, duplicate records and contradictory transaction treatments all need fixing before compliance can begin.
This is where the real value of agentic AI is emerging.
- Not as a replacement for tax professionals.
- Not as a replacement for tax engines.
But as a replacement for the endless manual effort spent trying to understand and repair poor-quality data.
Beyond the VAT chatbot
The winners in tax technology will not be those with the most impressive VAT chatbot.
A conversational interface is useful. Asking “Why did this transaction fail validation?” or “Prepare my Spanish VAT return” is undoubtedly easier than navigating multiple screens and reports.
But conversational AI only becomes valuable when it sits on top of a platform that can actually perform those tasks.
The real challenge remains underneath the conversation. Can the system understand a messy ERP export, identify what it contains, recognise relationships between fields, fill information gaps, validate transactions and produce data that is genuinely ready for compliance?
If not, the chatbot is simply providing a more attractive way to access bad data.
A mistake so bad, they’re repeating it in e-invoicing
The same applies to e-invoicing. The industry discussion often focuses on transmission networks, formats and government platforms. Yet many businesses are discovering that the biggest obstacle is not sending an invoice. It is ensuring the underlying transaction data is complete and correct before the invoice is created.
Where will AI land in the VAT automation architecture?
Agentic AI has the potential to sit between source systems and tax processes, continuously improving data quality, identifying anomalies and reducing the volume of manual intervention required from tax teams.
This is particularly important as tax authorities gain access to more granular transactional data through e-invoicing and digital reporting regimes – VIDA Digital Reporting Requirements. Errors that previously remained hidden until a VAT audit are increasingly visible in real time.
The future of AI in VAT is therefore unlikely to be defined by who has the smartest chatbot.
It will be defined by who can combine conversational AI with the ability to make sense of imperfect data.
Because in indirect tax, the biggest compliance risk is rarely the VAT return.
It’s the transaction data that feeds it.
VATCalc safely incorporating AI into tax calculations and returns
See how else VATCalc is applying Artificial Intelligence since 2023 to enhance VAT determination, returns and e-invoicing:
- AI and item classification in VAT determination;
- Tax advice generated by AI from our VAT expert-maintained tax engine
- Tax code mapping with AI to reduce significantly tax engine implementations and testing
We have since moved through to develop the world’s first Agentic AI VAT tax engine, embedding our tech across all platforms.
