AI Transparency Statement
Last updated July 15, 2026
AIOS builds AI agents that talk to people by voice and text. We believe a person should always know when they are talking to AI. This statement explains what our AI does and how we disclose it, in line with the EU AI Act (Article 50, which applies from 2 August 2026) and US AI-disclosure laws. AIOS is operated by NoManagement B.V. (Netherlands).
Why this statement exists
AIOS builds and runs AI agents that talk to people by voice and by text. We believe a person should always know when they are talking to AI. This statement explains what our AI does, how we disclose it, and how that maps to the EU AI Act's transparency rules (Article 50) and similar US AI-disclosure laws. AIOS is operated by NoManagement B.V. (KVK 80968724), Spijkerkade 31, 1021 JS Amsterdam, Netherlands (European Union).
On a phone call, the AI tells you first, in code
On a phone call, before any conversation begins, the agent speaks a set disclosure that it is an AI assistant calling from the business and that the call is recorded. This opener is enforced in our code: it runs before the conversation, the agent cannot skip or paraphrase it, and an un-removable default is always present, so the disclosure cannot be quietly switched off. Each call's disclosure is logged with its exact wording, a SHA-256 hash and a timestamp, giving a tamper-evident record that it was given. This directly serves EU AI Act Article 50(1), which requires that people be informed when they interact with an AI system unless that is obvious from the context, and Article 50(5), which requires the disclosure to be clear and given no later than the first interaction. These transparency obligations apply across the EU from 2 August 2026.
In text and web chat, the agent never poses as human
In text and web chat, the agent is instructed never to claim to be a person and to answer truthfully if it is asked whether it is AI. On our own AIOS website, the chat widget carries a visible "AI Agent" label. Across the text channels a business runs itself, such as SMS, WhatsApp, Instagram and embedded chat, there is no single system-forced first message, so making the AI disclosure clear on those channels is a contractual duty on the business that deploys the agent, as set out in our Acceptable Use Policy and Compliance Addendum.
Voice, cloned voices and deepfake transparency
AIOS speaks with a synthetic, computer-generated voice, and we disclose that it is AI up front on every call so a listener is never misled into thinking it is human. The AI Act treats AI-generated voice in two ways. Article 50(2) asks the provider of the AI system to mark synthetic audio so it is detectable as artificially generated, as far as the state of the art allows; we record the artificial origin of the audio in the call's data. Article 50(4), read with the definition of a deepfake in Article 3(60), asks the deploying business to disclose when a voice is generated to resemble a specific, identifiable real person. Our standard voices do not imitate a named real person; where a business chooses a voice cloned from a real person, that business must add the extra deepfake disclosure, and our spoken AI disclosure is given either way.
What the AI does and does not decide
AIOS agents answer questions, qualify leads, book appointments, send follow-ups and route conversations, using the knowledge and instructions the business gives them. They are assistive tools operating under the business's direction. They do not make legal, medical, financial or other professional decisions, and their output is not professional advice. AI can make mistakes, mishear or mis-transcribe, and we do not guarantee any particular outcome.
Human oversight
The business that deploys an AIOS agent supervises it: it sets the agent's instructions and knowledge, can review conversations and recordings, and can hand a conversation to a human. We provide the controls; the deploying business is responsible for supervising how its agent is used and for stepping in where its industry or the law requires a human.
Our role and yours under the AI Act
As the operator of the AIOS platform we build the AI system and bake the spoken AI and recording disclosure into the phone channel in code. As our customer, you are the party deploying that agent to contact people, so you are also responsible for meeting your own AI-disclosure obligations in the places you operate, including on the text channels you run and any obligations that go beyond ours. AIOS applying EU AI-transparency expectations to its agents does not relieve you of your own obligations. See the Compliance Addendum and Terms of Service.
Data, recordings and your rights
Calls and chats are recorded and transcribed to operate the service. How that data is handled, how long it is kept, the legal bases we rely on, our sub-processors, and how EU-to-US transfers are safeguarded are set out in our Privacy Policy and Data Processing Addendum. If you interacted with an AIOS agent and want to exercise a data right, email the address below and we will route it to the business responsible for that agent.
Standards we align to
We align our transparency practices to the EU AI Act, in particular the Article 50 transparency obligations for AI systems that interact with people and for AI-generated audio, which apply across the EU from 2 August 2026, together with the GDPR. We also align to US AI-disclosure laws such as California's bot-disclosure law (SB 1001) and Utah's AI Policy Act (SB 149), and to the all-party recording-consent state laws covered in our Recording and AI-Voice Notice. We track these laws as they change and update this statement accordingly.
Questions, or a privacy or data request? Email support@aios.supply.