VRsion 1 AI uses artificial intelligence. This notice explains, in plain language, where and how — and where it does not.
AI works. Humans decide.
The AI system
| System | VRsion 1 AI — AI Sales Workspace |
| Operated and provided by | VRSION PRO SRL, Belgium |
| Underlying AI model | A large language model supplied by OpenAI (currently GPT-4o mini), accessed through OpenAI's API |
| Intended purpose | Helping business sales teams qualify business prospects and prepare outreach drafts, for review by a person |
| Who uses it | Employees and contractors of our business customers |
Where AI is used
| Feature | What the AI does |
|---|---|
| Qualification | Writes a plain-language explanation of how well a prospect company fits the customer's ideal customer profile: reasons, risks, a recommended offer, a likely buyer role and an outreach angle. |
| Outreach drafts | Writes a first draft of an outreach email, in the language set for the prospect or the workspace. |
| Follow-up drafts | Writes a draft follow-up, which goes through the same review as any other draft. |
The fit score itself is not produced by AI. It is calculated by fixed rules that compare the prospect's industry, country and size with the customer's ideal customer profiles; the same inputs always give the same score.
AI is used only when a person asks for it by clicking an action. Nothing runs in the background on its own.
What the AI does not do
- It does not send emails. Only a person can approve a draft, and sending requires a further, separate action.
- It does not approve its own work.
- It does not decide on its own whom to contact.
- It does not run campaigns, sequences or automatic follow-ups.
- It does not take decisions with legal or similarly significant effects on individuals, such as decisions about employment, credit or access to services.
- It does not guess a person's language from their name or country; languages are always set by a person.
Human oversight
- Every AI-drafted email is shown in an Approval Center with its reasoning, the prospect's details and the policy decision, and must be approved by an owner or admin of the workspace.
- Approval does not send the email. Queueing it for sending is a second, separate human action, and sending must first be enabled for the workspace by VRsion Pro.
- The product states on screen that AI generates the drafts and can be wrong.
- Each workspace has a daily AI budget and a daily email limit.
- Proposals, approvals, rejections, queue actions and outcomes are recorded and visible in the workspace's activity history.
Your responsibilities as a customer
AI output can be incomplete or wrong, and it reflects the quality of the information it is given. Before acting on it, review the qualification and its reasoning, the prospect's details, any commercial claims or prices, and the content, recipient and language of each email — and make sure you have a lawful basis to contact the recipient. You remain responsible for the emails you approve and send. See our Terms of Service.
Emails received by prospects
Outreach emails are drafted with AI assistance, then reviewed and approved — or rejected and regenerated — and sent by a person at the business named as the sender, which takes editorial responsibility for them. If you received such an email and have questions, contact that business; you can also write to us at ai@vrsion.pro.
Data and the AI provider
To generate a qualification or draft, the service sends the customer's business context and the relevant prospect information to OpenAI. Under OpenAI's API terms, this data is not used to train OpenAI's models by default, and OpenAI may keep it for up to 30 days to monitor abuse. We do not use customer data to train AI models. Details are in our Privacy Policy and on our Subprocessors page.
Governance summary
| Intended purpose | Assisting business sales teams to qualify business prospects and prepare outreach drafts, for review and decision by a person |
| Roles | VRSION PRO SRL develops VRsion 1 AI and makes it available under its own name. OpenAI provides the underlying general-purpose AI model. Our customers use the system in their own business. |
| Model and provider | OpenAI, GPT-4o mini, through OpenAI's API |
| Risk classification | Based on its intended purpose, VRsion 1 AI is not used for a practice prohibited by the EU Artificial Intelligence Act (Regulation (EU) 2024/1689) and does not fall within the high-risk use cases listed in Annex III of the Act. We reassess this whenever the intended purpose changes. |
| Transparency | The product identifies AI-generated content to its users, and this notice describes how AI is used. We are assessing, in light of the evolving guidance on Article 50 of the Act, whether AI-generated text should also carry a machine-readable marking. |
| Evidence and logs | Every AI-generated proposal, policy decision, approval, rejection, queue action and sending outcome is recorded. |
| Human oversight | Approval required for every outreach email; approval and sending are separate human actions; sending must be enabled per workspace; daily limits and budgets. |
| Incidents | Reports to ai@vrsion.pro are reviewed and handled under our internal incident procedure. |
| Updates | Any change of AI provider or model is authorized internally and reflected in this notice and on our Subprocessors page before it takes effect. |
Reporting a problem
If you notice AI output that is harmful, discriminatory, or seriously wrong, tell us at ai@vrsion.pro. We review every report, and we record and investigate incidents under our internal incident procedure.
Updates
We update this notice when we change how AI is used, including a change of AI provider or model. The version and date at the top show the current version.