AI oversight your works council can say yes to.
Any tool that could monitor employees needs your works council on board. Alesia makes that conversation short: it acts only when something risky is about to reach an AI tool, and it never reads, stores, or reports what people write.
The boundary is structural. The prompt and the verdict stay on the device. The organization receives only category, severity, time, and whether the employee proceeded. That is a design you can write into a works agreement, not a promise you have to defend.
- Not a keylogger, tracker, or productivity monitor
- Acts only at the moment of risk
- Employees can override warnings
- Organization sees metadata, never content
Where the line actually sits
In Germany, section 87(1) no. 6 of the Works Constitution Act (BetrVG) gives the works council a say over any technical system that can monitor employee behavior or performance. Alesia keeps that surface small. It evaluates text only at the moment it's about to be sent to an AI tool. It doesn't read documents in the background, record keystrokes, or log browsing history. When a prompt matches a risk category, the check happens on the device, the employee sees the warning, and only the outcome (what kind of risk, how severe, whether they proceeded) reaches the organization. The prompt is discarded on the device once the check completes and is never stored anywhere.
The questions the works council asks
IT presents Alesia to the works council. The first question is the usual one: can management see what employees type? The answer is in the architecture, not in a policy document: no. Prompts are checked and discarded on the device. The second question: can a report single someone out? The dashboard shows events in aggregate, by category and severity, and overrides aren't tied to a person by default. The third: what happens with a false positive? The employee overrides the warning and keeps working, and their manager isn't notified. What's left to agree on is short: which risk categories to check, and how long metadata is kept.
What changes for your team
Rolling out AI oversight stops being a negotiation about surveillance. The architecture is the guarantee, it's the same on every device, and it's simple enough to describe in a works agreement. Employees keep working the way they do, with a warning only when something risky is about to leave.
- No prompt or chat content ever reaches the dashboard
- Overrides visible in aggregate, not tied to a person by default
- Short, concrete answers to the works council's questions
- Employees keep their way of working, without feeling watched
Say yes to AI.
Set the rules once. Alesia enforces them on every device, and prompts stay on your devices.