ZeeDo Acceptable Use Policy

1. Scope

1.1 This policy forms part of the ZeeDo Terms of Service. It applies to the Customer, its Members, and every bot the Customer runs on ZeeDo, on every plan.

1.2 The Customer is responsible for its Members and its bots under this policy as for its own conduct. "The bot did it" is not a defence, because a bot acts on the Customer's Instructions (Terms of Service, section 3).

1.3 Terms defined in the Terms of Service have the same meaning here. "You" means the Customer and each of its Members.

2. Access controls and websites

You must not use ZeeDo, or instruct or configure a bot, to:

2.1 defeat, bypass or outsource the solving of a CAPTCHA or any other human-verification challenge, or work around the hand-off by which a bot leaves such a challenge to a person;

2.2 circumvent an access control, paywall, rate limit, IP block, or technical measure a website or service uses to restrict access;

2.3 access a website or collect data from it in breach of that website's terms, including by scraping a site whose terms forbid automated access;

2.4 try to reach a site ZeeDo's site registry marks prohibited, for example by another domain name, a mirror, a redirect, or a browser a bot starts for itself;

2.5 falsely confirm that your organisation holds an account on a site, or the right to let a bot use it, including under the site's own terms;

2.6 disguise a bot's traffic as not coming from ZeeDo, or route a bot's traffic around ZeeDo's egress controls;

2.7 sign in to an account you are not entitled to use, or use credentials you are not authorised to use.

3. Regulated output, payments, commitments, automated decisions and high-risk uses

You must not:

3.1 rely on bot output in a tax, legal, medical or financial matter before you have checked it, or, where the matter calls for a qualified person, before one has reviewed it; or present bot output to others as professional advice;

3.2 try to make a bot pay, enter or read a payment instrument, commit funds, or submit a regulated filing, including by splitting the act into steps, by routing it through another tool, or by storing a payment card's details in the vault;

3.3 take, through a bot, a decision based solely on automated processing that produces legal effects concerning a person or similarly significantly affects them, unless a person with the authority and competence to change the decision reviews it meaningfully first;

3.4 use ZeeDo for a purpose that makes an AI system high-risk under Article 6 (2) and Annex III of the EU AI Act, for example filtering job applications or evaluating candidates or employees, assessing creditworthiness, pricing life or health insurance, deciding on access to education or to essential public services, or identifying people by biometric data. ZeeDo is intended for automating a business's office and online work, and is not intended or designed for such a use. A Customer that uses ZeeDo for one acts outside ZeeDo's intended purpose and is the provider of that high-risk AI system under Article 25 (1) (c) of the AI Act, with all of a provider's duties;

3.5 use bots to provide to others legal, tax or other services that the law reserves to authorised persons, unless you hold that authorisation and a qualified person reviews the work.

4. Fair use (every plan)

4.1 The rule

ZeeDo applies this rule on every plan:

"a workspace is flagged when its machine averages 90% or more of its allotted vCPU across a rolling 6-hour window with zero token and tool cost rows in that window. A flag notifies the workspace owner and our abuse queue; nothing is suspended automatically. Suspension requires a written notice, a 72-hour cure period and a human decision"

4.2 What this means in practice

• Every ZeeDo plan, Free included, comes with unlimited working hours. In return, workloads must serve the Customer's own business.
• Using a workspace's machine as general-purpose compute that has nothing to do with running bots is not fair use. Examples: cryptocurrency mining, media-transcoding farms, and general free hosting.
• The signal ZeeDo watches is simple: the machine is busy (90% or more of its vCPU, averaged over 6 hours) but its bots make no model calls and no tool calls in that time.
• A flag is a notification, not a penalty. It is shown in the workspace as soon as it is raised, at the same time as it reaches ZeeDo's abuse team.
• ZeeDo will not suspend a workspace for fair use unless it has:
  1. sent a written notice to the workspace Owner explaining the concern;
  2. allowed 72 hours for the Customer to fix it or explain; and
  3. had a person at ZeeDo decide that suspension is justified.
• Nothing in this section is enforced automatically. Fair use is never a ground for the immediate restriction in section 8.

5. Communications with third parties

You must not use a bot to:

5.1 remove, alter, hide or contradict the AI disclosure ZeeDo adds to bot messages, or claim that a bot is a human;

5.2 impersonate a person or organisation, or send messages that mislead recipients about who is sending them;

5.3 send unsolicited marketing, bulk messages or spam in breach of applicable law, including without the consent the law requires for advertising by email or message, or contact people who have opted out;

5.4 harass, threaten, defame or intimidate anyone.

6. The platform itself

You must not:

6.1 try to escape a bot's machine, reach another customer's workspace, or read another customer's data;

6.2 try to extract, read or exfiltrate credentials from the vault, other than through the audited reveal in the dashboard;

6.3 try to bypass the egress broker, the approval gates, the policy engine or the site registry;

6.4 test the security of ZeeDo except under the Security Disclosure Policy and within its scope;

6.5 overload the service, or run denial-of-service traffic from or against it;

6.6 resell the service or give access to it to anyone outside your organisation, except under an Enterprise order form that allows it;

6.7 create more than one Free organisation for the same business to get around Free-plan limits.

Security research that follows the Security Disclosure Policy does not breach this policy.

7. Unlawful and harmful content and activity

You must not use ZeeDo to:

7.1 create, store or send illegal content, including child sexual abuse material;

7.2 build, distribute or run malware, or attack systems you are not authorised to test;

7.3 infringe intellectual-property rights, privacy rights or other rights of others;

7.4 process personal data without a lawful basis, or in breach of data-protection law;

7.5 carry out a practice prohibited by Article 5 of the EU AI Act, for example manipulating people by subliminal or deceptive techniques, exploiting people's vulnerabilities, social scoring, recognising emotions at the workplace or in education, or building facial recognition databases by untargeted scraping of facial images;

7.6 break sanctions or export-control law, or use the service from a territory subject to comprehensive sanctions, or for the benefit of a person subject to sanctions, of the European Union, the United Nations or Germany or, as far as Council Regulation (EC) No 2271/96 and section 7 of the German Foreign Trade and Payments Ordinance (Außenwirtschaftsverordnung) allow, of the United States.

8. Abuse and security restrictions

8.1 ZeeDo signs each request a bot sends to a website as coming from ZeeDo, obeys robots.txt for the pages a bot chooses to open itself (except in an account the Customer holds on that site), and limits the rate of requests to each website. Websites judge ZeeDo's traffic as a whole, so one customer's abuse can cost every customer access.

8.2 For the cases in the Terms of Service, section 10.2 (illegal activity, attacks, malware, child sexual abuse material, sanctions, and threats to the platform or to third parties, including abusive traffic), ZeeDo may restrict or suspend a bot, a workspace or an organisation immediately, and may do so automatically. ZeeDo's egress broker runs an automated abuse detector over the requests a bot chooses to send. When it recognises scanning of many addresses, repeated failed sign-ins on one website, a known attack pattern or a download of known malware, it may block the workspace's requests to that website, or suspend the workspace, without a person deciding first. A person at ZeeDo decides the other cases. ZeeDo acts diligently, objectively and proportionately, gives the Customer a statement of reasons (8.4), and lifts a restriction once it is no longer needed, unless ZeeDo ends the contract under the Terms of Service.

8.3 Appeal. The Customer may appeal a restriction under 8.2 to abuse@zeedo.cloud or, for a suspended workspace, with the appeal form in the product. A person at ZeeDo reviews the appeal afresh and answers within 5 business days; where another person at ZeeDo can review it, that person did not take the original decision. The Customer may also go to court; it need not appeal first.

8.4 Statement of reasons. For each restriction under 8.2 or 9.2, ZeeDo gives the Customer a statement of reasons. It says what was restricted and for how long, why, the facts ZeeDo relied on, whether automated means detected the cause or took the decision, and how the Customer can challenge it (8.3 and 9.2). ZeeDo sends it by email to the Owners: no later than when the restriction takes effect where a person at ZeeDo imposes it, and without undue delay where the abuse detector imposes it. When a workspace is suspended under 8.2 or 9.2, the product also shows in that workspace, from the moment the suspension is recorded, why it was suspended and since when, the facts ZeeDo relied on, whether ZeeDo's abuse detector or a person decided, and how to appeal. ZeeDo leaves out only what the law requires or allows it to withhold.

9. Enforcement, notices and appeals

9.1 ZeeDo may investigate a suspected breach, and may review the records and the Customer Content concerned as far as the investigation needs. Outside the cases in 8.2, ZeeDo's first step is to contact the workspace Owner.

9.2 Depending on how serious the breach is, ZeeDo may warn the Customer, stop a bot, restrict a feature, restrict or suspend a workspace, or end the contract under the Terms of Service. ZeeDo gives a statement of reasons for each such restriction as 8.4 describes. The Customer may contest it by writing to abuse@zeedo.cloud, where a person at ZeeDo reviews it, and may go to court. A breach of this policy may also make the Customer liable to ZeeDo, including under the indemnity in the Terms of Service, section 21.

9.3 Fair-use enforcement always follows section 4, never section 8.

9.4 Notices of illegal content. Anyone may report to abuse@zeedo.cloud content stored on ZeeDo, or sent by a ZeeDo bot, that they consider illegal. A notice should give:

• why the person considers the content illegal;
• where it is: an address, a message, a time and the bot's disclosure line;
• their name and email address (not needed for a report of the sexual abuse or sexual exploitation of children, or of child sexual abuse material); and
• a statement that they believe the report is accurate and complete.

ZeeDo confirms receipt to the email address given, decides on the notice in a timely, diligent, non-arbitrary and objective way, and tells the person who reported it its decision, whether it used automated means, and how to challenge the decision. Other abuse by a ZeeDo bot may be reported to the same address. Where ZeeDo becomes aware of information giving rise to a suspicion that a criminal offence threatening a person's life or safety has taken place, is taking place or is likely to take place, it informs the law enforcement or judicial authorities concerned.

9.5 Security vulnerabilities. Report them to security@zeedo.cloud under the Security Disclosure Policy.

9.6 Points of contact. Authorities of the EU Member States, the European Commission and the European Board for Digital Services may contact ZeeDo at contact@zeedo.cloud, in English or German. Users of the service may contact ZeeDo at support@zeedo.cloud or through the contact form at https://zeedo.cloud/contact, where a person answers.