Setup assistant
The user’s answers and a limited dashboard context.
Structure company, team, targets, financial and growth setup.
Trust · AI transparency
This notice identifies CEO Dashboard’s AI features, the information they use, their limits, and the controls around consequential work. It supports transparency under the EU AI Act but is not certification or a universal risk classification.
AI disclosure
AI surfaces are labelled as Addy, setup assistant, AI editing or AI advisors.
The user’s answers and a limited dashboard context.
Structure company, team, targets, financial and growth setup.
Prompts, recent conversation and relevant company context.
Answer questions, analyse information, draft updates and propose actions.
Document sections and semantic embeddings; selected passages may be returned to Addy.
Find relevant company knowledge and ground answers with citations.
Selected text or extracted attachment text when the user requests editing or analysis.
Summarise, rewrite, structure or save the requested material.
Audio deliberately submitted through the microphone control.
Turn the recording into text for the user’s message.
Public pages from a website supplied by the user.
Prepare suggested company context for the user to confirm or correct.
The decision question and the available evidence context.
Challenge assumptions and produce advisory analysis; advisors do not decide.
Data sent for processing
The current prompt and recent turns needed to continue the conversation.
Relevant KPIs, plans, tasks, people, decisions, connector summaries and other dashboard context.
Relevant document passages, page text, selected text and extracted file content when those features are used.
Audio only when transcription is requested, and public website text only when research is requested.
Requests pass through Lovable’s managed AI service to selected model providers. The application currently uses language, embedding and transcription model categories. Models may change as the service evolves. Do not submit passwords, API keys, unnecessary sensitive personal data, or material you are not authorised to process. The product blocks some obvious secrets in typed chat, but this is not a complete redaction system for documents, attachments or connected data.
Purpose and limitations
AI can summarise, compare, search, draft documents, suggest tasks and prepare structured proposals from the context available to it.
Outputs may be incomplete, outdated, inaccurate, biased or unsupported. Citations show retrieved context, not that a conclusion is legally or factually correct.
The service does not replace legal, financial, employment, medical or other qualified professional judgement.
Expanded customer-facing model/vendor disclosure, known limitations, evaluation results and change records are planned.
Human oversight
AI agents may be Responsible for bounded work but are not assigned the Accountable role. The authorised human remains responsible for the decision and its effects.
Where a member cannot decide, proposed changes can be sent to an authorised human for approval or rejection. In team mode, a proposer cannot approve their own consequential proposal.
AI tool inputs are structured and validated. Many actions first return a proposal; the user interface applies the resulting update rather than giving the model unrestricted system access.
Users should check source quality, assumptions, calculations, affected people and downstream impact before relying on or committing an output.
Traceability
Addy threads and messages can be saved with their author and time and shared with members of the same company.
Proposals can record their author, status, human decider and decision time. Committed decisions can retain stage, statement, human decider and available provenance.
Selected actions produce company activity records, and AI requests carry a correlation identifier for service diagnostics.
Administrators cannot yet set automatic AI-log retention periods or export and erase all AI interactions through one self-service control.
Risk and restrictions
A general CEO dashboard label does not classify every customer deployment. Purpose, affected people, data, integrations and decision impact determine the relevant duties.
Do not use outputs without qualified review for employment or worker management, credit, essential services, biometric uses, safety, legal rights or similarly consequential decisions.
Customers must not configure the service for unlawful manipulation, prohibited social scoring, prohibited biometric categorisation or other practices prohibited by applicable law.
A documented in-product workflow for classifying deployments against prohibited and high-risk categories is planned and is not currently an enforced technical gate.
Customers should train users, give required notices to affected people, set meaningful oversight and ensure reviewers have authority to question, override or stop the system.
Roles and timing
When Fotando supplies the AI-enabled product under its name, it may have provider obligations for that system. Upstream model providers have separate obligations for their models.
A customer using the system under its authority is typically the deployer and controls the business purpose, users, source data and whether outputs affect people.
The Act entered into force on 1 August 2024. Prohibited-practice and AI-literacy rules applied from 2 February 2025; GPAI and governance rules from 2 August 2025; most remaining provisions apply from 2 August 2026.
Certain high-risk systems embedded in regulated products have a later scheduled application date. Formal amendments and official guidance should always be checked.
AI concerns and complaints
Email trust@fotando.global to report harmful, discriminatory, privacy-invasive, misleading or undisclosed AI behavior, an incorrect consequential output, or a failure of human oversight. Include the company, approximate time, feature, impact, and any conversation or decision reference you can safely share.
We identify the affected feature, records, data flow and responsible people or providers.
Where appropriate, we can investigate access, correct product information, restrict a workflow or identify further mitigation.
We explain the outcome and next route. Depending on the issue, rights may also be exercised under data-protection, consumer, employment or other law.
Official sources
Privacy and AI contact
Fotando Solutions Sp. z o.o. · Uladzimir Ausianikau, CEO. Contact us about personal data, AI behavior, safeguards, or your deployment.
This page is product information, not legal advice or a certification. Your obligations depend on your organisation, data, integrations, and use case.