How to prevent unwanted AI actions in Exact Online

An invoice that gets created by mistake. A customer record overwritten with the wrong data. A credit invoice on the wrong debtor. These are the scenarios that keep you up at night when you consider connecting AI to your accounting.

And rightly so. But the solution is not to avoid AI, it is to set up AI properly. In the first article of this series we discussed why responsible AI use starts with conscious choices. This article goes a step further: how do you concretely prevent the AI from doing things you do not want?

Nothing happens without your permission

The most important protection is simple: the AI connector performs no write action without your explicit confirmation. No invoice created, no record changed, no entry deleted, unless you give your approval.

Here is how that works in practice:

1. You give an instruction. For example: "Create a sales invoice for customer Ecolux, 10 hours of consultancy at 150 euros."
2. The AI shows a preview. You see exactly what will be created: customer name, amount, description, VAT code. Everything laid out.
3. You confirm or correct. Is it right? Then you confirm and the invoice is created. Is something off? You say "no" or adjust the data. Nothing changes in your records.
4. Extra warnings for risks. For amounts above 10,000 euros you get an additional warning. For delete actions you get yet another confirmation.

This is not an optional setting. It is a fixed part of the connector that cannot be switched off. Every mutation, whether creating, changing or deleting, goes through this confirmation step.

Ask good questions, get better results

The AI does exactly what you ask. That sounds like an advantage, and it is. But it does mean the quality of your question directly affects the result.

Be specific. "Show invoices" can produce all sorts of things. "Show outstanding invoices from January 2026 for customer Joanknecht" gives you exactly what you are looking for. The more concrete your question, the smaller the chance of unwanted results.

Name the entity. The AI works with customers, invoices, general ledger accounts and items. If you say "look up Joanknecht", the connector knows you mean a customer and remembers that context. If you then ask "show outstanding invoices", the AI automatically filters on that customer.

Break complex actions into steps. Do you want to create a credit invoice? First ask to show the original invoice. Check the data. Only then ask to create the credit invoice. Step by step.

Look first, then act

The most powerful habit you can develop: always ask to show data first before you make changes. This sounds obvious, but it makes a big difference.

Suppose you want to make an address change for a customer. Do not immediately ask "Change the address of Bakkerij Janssen". First ask: "Show the details of Bakkerij Janssen." You then see the current address, the customer number and any outstanding invoices. Only after that do you submit the change.

De koppeling helpt je hierbij. Wanneer je een klant opzoekt en er is precies één match, dan onthoudt de AI die context. Vraag je vervolgens om facturen, dan worden automatisch de facturen van die klant getoond. Zijn er meerdere matches, bijvoorbeeld drie bedrijven met “Janssen” in de naam, dan vraagt de AI welke je bedoelt voordat er iets gebeurt.

The connector helps you with this. When you look up a customer and there is exactly one match, the AI remembers that context. If you then ask for invoices, the invoices of that customer are shown automatically. When there are multiple matches, for example three companies with “Janssen” in the name, the AI asks which one you mean before anything happens.

Everything is recorded

Every interaction with the connector is logged automatically. Which question was asked, which tool was called, which parameters were passed, and whether the action succeeded or failed. This is not an optional feature, the audit log always runs.

That gives you two benefits. First, you can find out exactly what happened afterwards. If a colleague asks “who created that invoice?”, the answer is in the log. Second, it works preventively: knowing that everything is recorded encourages careful use.

Do’s and don’ts

What if something does go wrong?

Suppose you accidentally created an invoice that should not have been created. The audit log shows exactly what happened, including all parameters. You can find the invoice through the AI (“show the last created invoice”) and then create a credit invoice, again with the confirmation step.

The connector prevents most mistakes through the confirmation process. But mistakes that are confirmed by the user themselves are human errors. The audit log ensures they are always traceable and recoverable.

Staying in control is a habit

Responsible AI use is not a technical configuration you set up once. It is a way of working. Asking specific questions, checking data before you change it, reading previews before you confirm. The connector enforces the most important steps. The rest is discipline that you build up with experience.

In the next article we take a closer look at data privacy: which data does the AI process exactly, where is it stored and what about the GDPR?

Ready to get started? Start your free trial →