Acceptable Use Policy

GatPilot is a service operated by TOPFED WEB S.R.L., a limited liability company registered in the Republic of Moldova under IDNO 1026023125543. The company's registered office is in Chișinău, Republic of Moldova.

This policy forms part of the GatPilot Terms of service. It applies to every account, every user and every assistant. A breach of it is a breach of the Terms.

The account owner is also responsible for what the users they invite do.

Words have the meaning given in the Terms: account, assistant, source, conversation, channel.

• Terms of service
• Terms of service: the words we use

You do not upload, paste or add as a source:

• illegal content, or content that incites hatred, violence or discrimination;
• content you do not have the right to use, for example other people's texts, images or trademarks;
• files containing malicious software or designed to damage the service;
• other people's personal data, without a legal basis;
• data about health, religion, political opinions and the other special categories of data, without a legal basis and without telling us;
• other companies' trade secrets;
• passwords, access keys or card details.

The assistant answers from everything you have given it, including prices, discounts and stock. That is why you are responsible for keeping this content correct and up to date.

You write the assistant's name, the greeting and the “Personality and rules” field yourself, in the dashboard. You do not set it up:

• to present itself as a person or to deny being an AI when asked;
• to state prices, stock, discounts or conditions that you do not offer;
• to pass itself off as another company or another person;
• to ask for passwords, card details or other sensitive data;
• to use techniques that exploit someone's age, disability or vulnerable situation;
• to give diagnoses, or legal or financial advice, presented as definitive;
• to incite hatred, violence or discrimination.

Our rules require the assistant to answer only from its sources and not to invent prices or offers. You do not get round them through what you write in “Personality and rules”.

GatPilot itself says that an AI assistant is replying: in the chat window and at the start of the conversation on Telegram and Viber. You do not contradict this through the greeting, the assistant's name or “Personality and rules”.

• AI transparency

You may not:

• send spam or unsolicited bulk messages, through GatPilot or through the bots connected to it;
• break the rules of Telegram, Viber or any other connected platform;
• add people as team members or as report recipients without a legal basis or without informing them;
• collect data about people through the assistant without telling them why and for how long you keep it.

The assistants answer only the people who write to them. The only messages sent without a question are the reports, which go only to people you have added and who have activated their link.

GatPilotBot reads a website when you add it as a source. It does not check whether the website is yours, which is why you are responsible for making sure you may use it.

The bot reads robots.txt at the start of every crawl and pauses between batches of requests. The GatPilotBot page explains how it reads the rules in it.

You may not:

• add as a source websites you do not have the right to use;
• try to get round the rules in robots.txt or the bot's limits;
• point the bot at addresses on internal or private networks.

• GatPilotBot

You may not:

• create multiple accounts to get round the plan's limits or a suspension;
• use AI keys that do not belong to your company;
• share sign-in details or let people outside the company use the account;
• get round the request limits or overload the service;
• hide the “Powered by GatPilot” line in the widget by other means, on the plans where it stays.

Each plan's limits are in the Terms, in section 6.

• Terms of service

You may not:

• look for vulnerabilities in our systems without our written consent; you report them to us at info@gatpilot.com;
• reverse-engineer the service, except where the law allows it;
• try to reach the data of other GatPilot customers;
• put hidden instructions in pages, documents or messages, intended to manipulate other companies' assistants;
• get round the list of authorised websites for the widget or use another company's widget.

The assistants' rules require them to treat the text of the sources as material to read, not as orders. This reduces the risk, but it is not a guarantee.

Without a separate written agreement, you do not use GatPilot for decisions on hiring or on evaluating employees. The same applies to credit, insurance, access to essential services, education or law enforcement.

The rule also applies to other uses classified as high-risk by the EU Artificial Intelligence Act.

You do not use GatPilot at all for the practices prohibited by the same Act. For example, manipulation that exploits people's age, a disability or their vulnerable situation.

If an assistant built on GatPilot answers with illegal, misleading or harmful content, write to us at info@gatpilot.com. Send the website address or the bot's name, the date and a screenshot.

If GatPilotBot is putting too much load on your website, send the domain and the time. You also write to the same address about vulnerabilities found in GatPilot.

We reply within one business day, in English, Romanian or Russian.

We review every report and may ask you for information. We may stop an assistant or a channel and delete the content concerned.

We may suspend or close the account and notify the authorities, when the law requires or allows it.

We tell you the reason, unless the law prevents us or doing so would compromise an investigation. Where the risk allows, we tell you beforehand and give you time to put things right.