AI tools at work · Written by Yobel Tzegai · Last checked 8 November 2026
Salesforce AI under the EU AI Act: you are the deployer of what you switch on; lead scoring is usually not high-risk, scoring people for credit is.
Get my risk classes, freeBook a 20-min intro call20 minutes with the founder.
| Use | Usual outcome | Provision | What would change it |
|---|---|---|---|
| Scoring company leads, drafting emails, forecasting | Not high-risk | Article 6(2) and Annex III, which list none of them | Scoring natural persons for credit, point 5(b) of Annex III |
| An agent or chatbot that talks to customers | A transparency duty for its provider | Article 50(1) | Deep fakes it generates that you publish, Article 50(4) |
| Evaluating a person’s creditworthiness | High-risk | Point 5(b) of Annex III | A use solely for detecting financial fraud, which point 5(b) excepts |
| Ranking job applicants in a recruiting app | High-risk | Point 4(a) of Annex III | An Article 6(3) assessment, documented under Article 6(4) |
Add the AI you use. Each system gets a role, a risk class and the article it rests on.
LiveLink each duty to a file and the page it stands on: a policy, a log setting, a training record.
LiveAnswering a customer's questionnaire inside the app, from the proof you already linked.
LiveRules decide. AI only drafts. A person confirms.
€499 a month
or €4,990 a year. Prices exclude VAT.
Not as such. Annex III does not list scoring sales leads or forecasting revenue. It lists evaluating the creditworthiness of natural persons or establishing their credit score, in point 5(b), with an exception for detecting financial fraud. A score about a company is not a score about a natural person.
Not your part. The provider carries the provider’s duties. For a high-risk use you carry the deployer’s under Article 26: using the system as instructed, assigning human oversight, monitoring it, keeping its logs and informing the people affected. Ask for the instructions for use, and record each use in your inventory.