Use of artificial intelligence

OFICIA’s virtual employees use artificial intelligence to prepare work. For example, draft emails, payment reminders, replies to customers, reports and analyses.

They work with what your team asks for and with the data your company has saved in OFICIA.

Today they don’t send anything or change anything in your programs. They prepare the work, and a person decides what to do with it.

What the AI prepares is a proposal. Someone on your team reviews it, corrects it if needed and decides whether to use it.

Tasks that need it go through approval: a manager approves or rejects them, and the app records who did it and when.

The AI makes no decisions with legal effects on you or on anyone else.

The European Artificial Intelligence Act requires telling people when they are interacting with an AI. This is its article 50, which applies from 2 August 2026.

In OFICIA, every virtual employee carries the “AI” label next to its name.

If someone asks, it says it is an AI. It never pretends to be a person or signs with its own name: it leaves a space for whoever sends the text.

Virtual employees have a job title and may have a name, such as Elena. Your company can change it. They are not real people.

In the app, the AI’s answers come with the notice “Answers are generated by AI. Review them before you use them.”

If you copy a draft to use outside OFICIA, the label doesn’t travel with the text. Someone on your team reviews it and decides how to present it.

If you publish a text prepared with AI, bear in mind that in some cases the law requires you to say so.

When you make a request, we send the AI provider what it needs to answer:

• Your request and the latest messages in that conversation.
• The virtual employee’s instructions and a summary of its recent work.
• Your company profile and the information your team has saved for the virtual employees.

We don’t add your name, your email or your payment details.

We don’t use your content to train AI models.

With Google Gemini we only use the paid service. Under its terms, Google doesn’t use those requests to improve its products. We also ask it not to keep them in its request logs. Google keeps them for 55 days only to detect abuse.

See the list of providers

Today we use this one:

• Gemini, by Google.

We review the models often and switch when another one works better. If a provider changes, you will see it here and in the privacy policy.

AI can get things wrong. It may invent a fact, misunderstand a request or use outdated information.

It only knows what you tell it and what your company has saved in OFICIA. It doesn’t look into your programs.

It doesn’t replace your tax, employment or legal adviser.

That’s why everything it prepares goes through a person before it is used.

If you see a wrong or inappropriate answer, write to us at [LEGAL_EMAIL].