AI transparency notice

techtropy builds the features described here on third parties' general purpose models. We call them through OpenRouter Inc., New York, USA. In the language of the AI Act we are the provider of these systems, because we put them into service under our own name. For Intelligence we are also the deployer, because we run it on our own data.

We are not the provider of the underlying models. None of these systems is a high-risk system under Annex III, and none serves a practice prohibited by Article 5.

Strategy reads your uploads and drafts strategies from them, and it grades a strategy against each principle with a written reason. Both the draft and the grade come from a model.

Intelligence drafts entries in the company register from public sources. Companies, products, claims, relationships and job postings all reach the register as drafts.

In the simulation the strategy fox comments on a company's strategy and on the decision sheet it is filling in. It does so only when the executive asks it to.

On the home page the strategy fox answers questions on strategic management and on our products. It answers only from sources our staff choose, and names them under each answer. It declines questions those sources do not cover.

A simulation's market is computed by fixed rules that people wrote, with no model and no randomness. A period computed twice gives the same result. The computer companies follow hand-tuned rules.

The only model in the simulation is the strategy fox. It writes comments, gives no grade, rank or verdict, reports nothing to the instructor and affects no result. Nobody in a seminar is graded by a model.

Under Article 6 (3) of the AI Act it is not a high-risk system. It improves the result of an activity people completed before, namely the strategy. It decides nothing about people.

Where a model wrote something you read, the app says so at that very spot. The record carries the same mark. This way it survives an export or a later reading. A model can be wrong and sound confident doing so. Generated text is therefore a draft to check, not a source.

Generated text is a draft and never an outcome. In Strategy your team edits every field the model filled, and the grade informs your own judgment instead of replacing it.

In Intelligence a person reviews every draft that names a person before it becomes visible. Other drafts are reviewed by a person when the model states its reliability below a threshold we set. Drafts above the threshold become visible without prior review and carry the mark as generated. Our team corrects every reported entry.

On the home page our staff read the strategy fox's answers and correct the sources it answers from.

No decision about a person is made by a model. There is no profiling of individuals, no rating or ranking of people, no emotion recognition and no biometric processing anywhere in the Service. We make no decision based solely on automated processing within the meaning of Article 22 GDPR, which the privacy notice states as well.

The strategy fox on the home page sends your question, the conversation before it and the passages it answers from.

Strategy sends your uploads and what you write in the editor. The simulation's strategy fox sends a company's strategy and decision sheet. Intelligence sends public pages it fetched. Each call goes through OpenRouter Inc., New York, USA, to a provider that stores nothing and may keep nothing for training.

The privacy notice governs what happens to personal data, including where it is processed.

Models currently in use: …. We may change models and instructions, and the rules on this page apply to every model.

Write to privacy@techtropy.com about anything on this page, including a generated entry you believe is wrong. We correct the entry and we say what changed.