Product terms & conditionsEffective September 8, 2026Version 3.1
AI Features Terms
These AI Features Terms govern the parts of the WorkMonitor Services that generate written summaries, digests, narratives and answers using large language models ("AI Features"). They form part of the Platform Terms of Service, and capitalized terms have the meanings given there.
Two commitments sit at the top of this document because they are the ones that matter: we do not train models on your data, and no AI output may be the sole basis for a decision that materially affects a person.
We do not train on your data
1.1We will not use Customer Data — including activity records, screenshots, timesheets, messages, project data, or anything else you or your Monitored Users put into the Services — to train, fine-tune, pre-train, align, evaluate or otherwise develop or improve any generative artificial-intelligence or machine-learning model, whether ours or a third party's.
1.2We contract with every AI sub-processor we engage on terms that prohibit them from using the data we send them to train or improve their models, and that require them to retain it only for as long as is needed to return the result.
1.3We will not sell Customer Data to a model developer, licence it as training data, or make it available to a data broker, in identifiable or de-identified form.
1.4The aggregated, de-identified statistics described in the Platform Terms of Service are not used for model training either. The commitment above is not qualified by that section.
1.5If we ever intend to change this position, we will treat it as a material change to the Agreement: at least 30 days' notice, a positive opt-in from account administrators, and no application to data collected before you opted in. We do not currently intend to change it.
AI output is a suggestion, not a determination
2.1AI Features produce text that describes patterns in data. That text can be wrong. It can misread a quiet week, miss context the software never had, repeat a bias in the underlying signals, or state something with a confidence the evidence does not support.
2.2You will not use AI output as the sole basis for a decision that produces a legal effect for an individual, or a similarly significant effect on them. This includes any decision about discipline, dismissal, redundancy, pay, bonus, promotion, demotion, a formal performance rating, a probation outcome, or a report to a regulator or a client about an individual's conduct.
2.3Before such a decision, a person with the authority and the competence to reach a different conclusion must review the underlying data, must be able to disregard the AI output, and must have the information needed to do so. This obligation is yours as controller, and it exists whether or not the law applying to you uses the phrase "automated decision-making".
2.4You will tell your Monitored Users that AI-generated summaries about their work are produced, where the law requires it, and you will provide a route for a person to contest one. We provide the underlying data through the Services so that a contested summary can be checked against it.
What is sent, and when
3.1AI Features operate on a prompt that we construct from data already in your account. Depending on the feature, that prompt may include the name of a team member, their aggregated activity metrics for the period, the titles of applications and sites recorded in that period, project and task names, and the text of records your users have entered.
3.2Most AI Features run only when a person asks for one. Where your organization enables the scheduled summary agent, they run automatically on a weekday schedule for each member in scope.
3.3Where an AI provider is configured for your deployment, the prompt is sent to that provider for processing and the result is returned to your account. Where no AI provider is configured, every one of these features falls back to a summary computed entirely within our own infrastructure, and no data leaves it.
3.4The AI sub-processors we engage, what each is engaged to do, and where each processes data, are named in the Sub-processor list. Adding or replacing one is subject to the same notice and objection rights as any other sub-processor.
Screenshots and captured images
4.1We do not send captured screenshots to an AI provider. AI Features operate on metadata and on aggregated metrics, not on the images themselves.
4.2If that ever changes, it would be a material change to the Agreement and to the Sub-processor list, subject to notice, to your right to object, and to a positive opt-in.
Turning AI Features off
5.1AI Features that call an external provider operate only where an AI provider is configured for your deployment. Where one is not configured, those features fall back to processing inside our own infrastructure, as described above.
5.2Where you need a deployment in which no Customer Data reaches an external AI provider at all, tell us at legal@workmonitor.ai and we will confirm the configuration in writing before you rely on it.
Your responsibilities when using AI Features
6.1You will not use AI Features for a purpose prohibited by the Acceptable Use Policy, and in particular you will not use them to infer or record a special category of personal data about a person, or to profile a person for a purpose unrelated to their work.
6.2You will not submit to an AI Feature any data that you are not entitled to process, and you will not use AI Features to process data of a kind your legal basis does not cover.
6.3You are responsible for reviewing AI output before you rely on it, act on it, forward it, or put it in front of the person it describes.
6.4You will not represent AI output to a third party as a determination made by WorkMonitor, or as a finding we stand behind.
Ownership of output
7.1As between you and us, output generated for you by an AI Feature from your Customer Data is treated as Customer Data, and you own it on the same terms.
7.2Generative models can produce similar output for different users from similar inputs. We make no representation that output is unique to you, and we do not warrant that it does not resemble output produced for someone else.
7.3We claim no ownership of your output and will not use it to serve another customer.
No warranty for AI output
8.1AI Features are provided without any warranty as to accuracy, completeness, reliability, fitness for a particular purpose, or freedom from bias. The disclaimers in the Platform Terms of Service apply in full to them.
8.2We do not warrant that an AI Feature will produce the same output twice for the same input, or that an output will remain available or reproducible.
8.3AI Features are not a compliance control, an investigation tool, a legal assessment, a medical or psychological assessment, or a substitute for a manager's judgement, and must not be used as one.
Changes to AI Features
9.1We may change the models we use, the providers we engage, the prompts we construct and the way a feature behaves, in order to improve quality, reduce cost, or address a safety or legal issue.
9.2A change of AI sub-processor is notified under the Sub-processor list. A change that materially reduces a core AI capability you are paying for is notified under the Platform Terms of Service.
Where these terms sit
10.1If these AI Features Terms conflict with the Platform Terms of Service on a subject specific to AI Features, these terms apply to that subject. On every other subject, the Platform Terms of Service apply, and the Data Processing Addendum governs any conflict about the processing of Personal Data.