Last updated: August 28, 2026
How we classify Botchi under the AI Act, the controls the product provides, and the responsibilities that remain with customers.
This page describes Botchi's current position under Regulation (EU) 2024/1689 on artificial intelligence, as amended by Regulation (EU) 2026/1744. It is based on the consolidated AI Act and guidance published by the European Commission.
This is not a certification, an EU declaration of conformity, or legal advice. The AI Act assigns obligations based on role, intended purpose, and actual use. No public document can certify every configuration a customer may create.
Botchi offers an AI assistant, agent, and workflow system under its own name. Botchi therefore normally acts as the provider of the AI system. It integrates third-party general-purpose models and acts as a downstream provider in the AI value chain.
Botchi does not currently train or place its own general-purpose AI model on the market. The obligations for providers of general-purpose AI models under Articles 51-55 therefore normally remain with the model providers. Botchi must still understand the capabilities, limitations, and conditions of the models it integrates.
A business using Botchi under its authority normally acts as a deployer. A customer may take on provider obligations if it sells the system under its own name, substantially modifies it, or changes its intended purpose, as explained in Article 25.
Botchi is a general-purpose platform. It is not automatically a high-risk system merely because it uses generative models or agents. Classification depends on the specific function and intended purpose.
The standard Botchi service has not completed a conformity assessment for high-risk uses under Annex III or regulated products under Annex I. Before using Botchi for recruitment, worker evaluation or management, education admission or assessment, credit, life and health insurance, access to essential services, biometrics, critical infrastructure, law enforcement, migration, or the administration of justice, a customer must request a specific assessment. Human review alone does not make a use non-high-risk.
Article 6 contains limited exceptions for some procedural, preparatory, or supporting tasks that do not materially influence a decision. An operator relying on an exception must document its reasoning. A system that profiles people remains high-risk in the covered Annex III cases.
| Date | Main rules |
|---|---|
| February 2, 2025 | Definitions, AI literacy, and the main prohibited practices. |
| August 2, 2025 | Governance and obligations for providers of general-purpose AI models. |
| August 2, 2026 | Most remaining provisions, including Article 50 transparency, and enforcement powers. |
| December 2, 2026 | Transitional deadline for Article 50(2) marking for certain generative systems already on the market before August 2, 2026. |
| December 2, 2027 | High-risk requirements for Annex III systems. |
| August 2, 2028 | High-risk requirements for systems connected to Annex I products. |
Article 4 requires providers and deployers to take suitable measures supporting AI literacy for staff and people operating AI systems on their behalf. Effective August 2, 2026, Botchi adopted an internal AI literacy policy and a policy for prohibited uses and uses that require review. The responsible role is Founder / AI compliance owner. Article 5 prohibits practices including certain forms of harmful manipulation, exploitation of vulnerabilities, social scoring, and specified biometric uses.
Botchi presents the product as an AI assistant and agent system. Article 50 also requires a person to know, no later than the first interaction, that they are interacting with an AI system unless that fact is obvious from the context.
Providers of systems that generate text, images, audio, or video must make outputs detectable as artificial through machine-readable marking, as far as technically feasible. Botchi has completed a technical audit of its current flows. It preserves markings supplied by upstream models or services when the file format and export flow allow it and has documented the limits affecting free-form text and artifacts composed by Botchi. Not every exported format currently carries a persistent marking. For the system already on the market before August 2, 2026, Botchi keeps this work in its compliance programme for the December 2, 2026 transitional deadline and monitors solutions offered by providers and technical standards.
Anyone publishing deepfakes must disclose their artificial nature. Anyone publishing AI-generated text on matters of public interest must provide the required disclosure unless the content has undergone human review with editorial responsibility. The Commission's Article 50 transparency guidelines explain these duties.
These controls help govern a deployment. They do not replace risk classification, internal instructions, training, monitoring, or a conformity assessment when one is required.
The customer must define the intended use, appoint competent people for oversight, train users, configure access and approvals, monitor runs, and comply with rules that apply to its sector. The customer must also assess the GDPR, employment law, copyright, consumer protection, and contractual obligations. The AI Act does not replace those rules.
Customers must not present outputs as certainly correct or delegate a regulated decision to Botchi without first assessing classification, legal basis, data quality, oversight, and the ability to challenge the result. Contact Botchi before putting a possible high-risk use into production.
Botchi is subject to the general and transparency obligations that apply to AI systems. The standard service has not undergone a conformity assessment for high-risk uses.
The applicable obligations depend on the specific function and intended purpose. Before using Botchi for recruitment, worker evaluation, credit, education, or access to essential services, the use must be classified and, where required, the controls for high-risk systems must be adopted.
Botchi has adopted internal controls for user transparency, AI literacy, prohibited practices, and uses that require a specific assessment. It has also completed the technical audit of marking in its current flows and continues to monitor providers, formats, and standards toward the applicable deadline. This page describes the product's current position and is not a certification.
For questions about the AI Act, a possible high-risk use, or an AI system incident, email hello@botchi.ai with the workspace, affected feature, and a short description.