AI Transparency Notice — Anna · imanna.ai
INFINITESUM believes the use of artificial intelligence should be transparent. This Notice explains, in plain language, what Anna is, how it works, its limitations, and your responsibilities when using it. It supplements the Terms of Service and the Privacy Policy.
1. You are interacting with artificial intelligence
Anna is a generative artificial intelligence system. The answers, code, text, and documents it produces are generated automatically by AI models, not written by a human professional. Whenever asked, we will make clear that you are interacting with an AI assistant.
2. How Anna works
Anna takes your instructions and the necessary context and uses third-party AI models (for example, Anthropic and OpenAI) to generate an output. Details on data processing are in the Privacy Policy.
3. Limitations you should know
- AI can be wrong. Generative models produce inaccurate, outdated, or incorrect information, and may "make up" seemingly plausible content — known as "hallucination".
- AI can be wrong about itself. It may state that it cannot do something it can, or that it did something it did not. When it says it performed an action — created, sent, scheduled, fixed — verify the result. The product provides screens for that check (Projects, Usage, Scheduling), and we recommend using them.
- Generated code may contain flaws. Logic errors, vulnerabilities, or incompatibilities. Review, test, and validate before use or production deployment.
- Generated documents may contain content and formatting errors. See Section 5.
- Cost estimates are estimates. See Section 4.
- Bias may occur, reflecting the models' training data.
- Not professional advice. Anna's responses do not constitute legal, accounting, tax, financial, medical, or security advice. For sensitive decisions, consult a qualified professional.
4. Cost estimates
Before running work that consumes Credits, Anna presents an estimate and requires your approval. That estimate is generated automatically from a reading of your request and may fall above or below the actual cost. You pay what is actually consumed, and spend by project is recorded on the Usage screen.
5. Generated documents
Anna generates documents for your use — including draft contracts, quotation templates, and tax documents. These are automatic templates: they may contain content, calculation, or formatting errors, or clauses unsuited to your specific case.
Before signing, sending to third parties, or relying on a document generated by Anna, review it in full and, where applicable, have it reviewed by a qualified professional. Responsibility for the document you choose to use is yours.
6. When Anna talks to your customers
If you activate the Customer Assistant Module, Anna will answer your end customers on your behalf. In that case:
- you must inform your customers that support is provided by artificial intelligence, and offer a path to a human — this is required by messaging platforms and is regulatory good practice;
- you are the controller of those customers' data, and INFINITESUM is the processor, under the DPA;
- the answers given on your behalf are your responsibility toward your customer.
We provide standard disclosure text for this purpose.
7. When Anna acts on your calendar
If you connect the Scheduling Module, Anna may read your appointments, create events, and send invitations, on your behalf, to third parties you designate. You are responsible for that designation and for the contact details you provide.
8. Human oversight is essential
Anna is a tool to support your productivity — not a substitute for your judgment. The final decision and responsibility for using the output are always yours. We recommend human review of what Anna produces, especially in critical contexts: code going to production, documents going to third parties, messages going to customers, and appointments going into your calendar.
9. Your data and model training
As stated in the Terms of Service and the Privacy Policy, we do not use your project content to train AI models without your consent. Snippets are sent to AI providers only to perform the task you requested.
Data obtained from third-party integrations — for example, your connected calendar — is not used to train, improve, or personalize generalized AI models, as detailed in Annex I to the Terms of Service.
10. Responsible use
You agree not to use Anna for unlawful, deceptive, or harmful purposes, under the Acceptable Use clause of the Terms, including generating content that impersonates a real person or organization. Using AI does not remove your legal responsibility for the output you choose to rely on.
11. Automated decisions and human support
Some Service mechanisms are automated, including security controls (Credit freeze and investigation hold) and Anna's own responses. Under Article 20 of the LGPD, you may request a review of automated decisions affecting your interests and information about the criteria used, subject to trade and industrial secrets. You may also speak with a member of our team at support@imanna.ai.
12. How to report generated content
If Anna generates offensive, incorrect, misleading, or inappropriate content, report it. It helps us fix it and we take it seriously.
- In the application: through the support channel available on the Help screen
- By e-mail: support@imanna.ai
13. Contact
Questions about this Notice or the use of AI in the Service: support@imanna.ai. Personal-data matters: support@imanna.ai.
INFINITESUM LTDA · CNPJ 65.498.658/0001-10 · support@imanna.ai · support@imanna.ai · imanna.ai