How AI Workforce uses AI
Each AI employee is a set of instructions, tools and checks running on large language models, mainly Claude by Anthropic. The AI CEO plans the work, the AI employees carry it out with the tools you connect, and results are checked before they reach you. Our team sets up and supervises the system.
- You stay in control. Anything you want to see first waits in Approvals. You decide what may run on its own.
- Only what a task needs. We send the AI only the information needed for a task.
- No training on your data. We do not use your content to train AI models.
- Transparency. The dashboard shows which AI employee did what, and when.
Sara, our website assistant
- Sara answers questions on our public pages. She is an AI and says so.
- She answers from our help center and pricing information, does not make offers or promises, and passes anything she can't answer to a person.
- Her answers are general information. Your contract, prices and terms are only those in our Terms of service and your plan.
- She is not available inside the dashboard; signed-in customers send requests to our team there.
The setup assistant
The setup assistant asks about your business and turns your answers into a launch plan with tasks for each AI employee. It works with fixed rules you can see in the plan before you confirm it; nothing is published and nothing is connected until you agree and our team sets it up with you.
Know the limits
AI can be wrong, out of date, or produce content that resembles existing work. Check facts, numbers, legal statements and anything sensitive before it goes out. The legal AI employees flag common risks but do not give legal advice.
Labelling AI content
Some platforms and laws require you to label AI-generated or AI-edited images, videos or ads. We can add labels for you; you are responsible for choosing when they are needed for your audience and markets.
Not allowed
You may not use AI Workforce to:
- break the law or anyone's rights, including copyright, trademarks, privacy and data protection;
- send spam or unsolicited mass messages, or buy and scrape contact lists without a legal basis;
- deceive people, for example fake reviews or testimonials, impersonating real people or companies, or hiding that a message is advertising;
- create or spread disinformation, hate, harassment, threats or content that sexualises minors;
- promote weapons, illegal drugs, gambling without a licence, or financial schemes that mislead investors;
- create malware, attack systems, or get around security measures or platform limits;
- make decisions about people's employment, credit, housing or access to services without human review.
The usage policies of our AI provider and of the platforms you publish on also apply.
What happens if the rules are broken
We may refuse tasks, remove content from the workflow, pause the AI team or end the subscription, depending on how serious the breach is. Where possible we talk to you first.
Report a problem
If you see AI Workforce content that you believe breaks these rules or your rights, tell us at or through the support page.