NeboAI gives your business AI employees that do real work on real accounts: they read your mail, book your jobs, talk to your customers, and move money when you tell them to. That reach is the point, and it is why this policy exists. It sets out what you may not do with NeboAI, what we expect when an AI employee acts on someone's behalf, and what happens when the rules are broken. It applies to everyone who uses NeboAI, Nebo, the marketplace, Channels, the Janus AI gateway, and any AI employee, skill, plugin, or app distributed through them, and it is part of our Terms of Service.
1. The three rules everything else follows from
- Comply with the law. Wherever you and the people your AI employees contact are, the law that applies to a human employee doing the same job applies to the AI one.
- Do not cause harm. Not to the people you contact, not to third parties, not to the systems you connect.
- You are the employer. An AI employee acts under your instructions, on accounts you connected, with permissions you granted. What it does, you did.
2. Prohibited uses
You may not use NeboAI, or direct an AI employee, to:
Harm people
- Create, request, or distribute child sexual abuse material or any sexual content involving minors. We report this to the National Center for Missing and Exploited Children and to law enforcement.
- Threaten, harass, stalk, intimidate, or incite violence against anyone; or promote self-harm, eating disorders, or suicide.
- Provide instructions for weapons capable of mass casualties — biological, chemical, nuclear, radiological — or for attacks on critical infrastructure.
- Facilitate human trafficking, sexual exploitation, or the sale of illegal goods, drugs, or controlled substances.
Deceive or defraud
- Run scams, phishing, pyramid or advance-fee schemes, fake reviews, fake testimonials, or any fraudulent charge, refund, or chargeback scheme.
- Impersonate a person or organization, or present an AI employee as a human when a reasonable person would be misled and the law or the context requires disclosure.
- Create or spread disinformation, deepfakes of real people without consent, or coordinated inauthentic behavior.
- Manipulate elections, suppress voting, or misrepresent voting procedures.
Attack systems or violate privacy
- Write, deploy, or distribute malware, ransomware, or tools whose purpose is unauthorized access, denial of service, or evasion of security controls.
- Access accounts, systems, or data you are not authorized to access, or scrape, collect, or infer personal information about people without a lawful basis.
- Track people's location, communications, or activity without their knowledge and consent, including through stalkerware or covert monitoring of employees or family members beyond what the law allows.
- Bypass rate limits, authentication, quotas, or the permission model of NeboAI, Nebo, or any connected service.
Spam and abusive communication
- Send unsolicited bulk email, SMS, calls, or messages, or contact people who have opted out. Every outbound message an AI employee sends must comply with the rules for that channel — CAN-SPAM, TCPA, GDPR, and the terms of the service it goes through — including honest sender identity, working opt-outs, and calling-hour limits.
- Generate reviews, ratings, comments, or social engagement that misrepresent who wrote them or how many people hold a view.
- Use a connected account (Gmail, Google Workspace, Slack, a phone number, an ad platform) in a way that violates that provider's terms.
Discriminate
- Make or support decisions about employment, housing, credit, insurance, education, or access to goods and services on the basis of a protected characteristic, or in a way that has that effect where the law prohibits it.
- Produce content that dehumanizes people or promotes hatred based on who they are.
Misuse the platform
- Resell, sublicense, or wrap Janus AI gateway access, or use it to train a competing model, unless we have agreed to it in writing.
- Publish to the marketplace anything that hides what it does, requests permissions it does not need, exfiltrates data, contains malware, or misrepresents its author, pricing, or capabilities.
- Interfere with other customers' use of the Service or with our operation of it.
3. Uses that need extra care
Some jobs are lawful but consequential. If you put an AI employee on any of these, you must keep a person in the loop for decisions that affect someone's rights, money, health, or safety, tell affected people that automation is involved where the law or context requires it, and meet every professional and regulatory obligation that would apply to a human doing the job:
- Medical, legal, financial, tax, or insurance advice and decisions.
- Hiring, firing, evaluation, or compensation of people.
- Credit, lending, tenant screening, or eligibility determinations.
- Anything involving minors, patients, or other vulnerable people.
- Payments, refunds, transfers, and other movements of money.
- Emergency services, safety-critical systems, or physical equipment.
Nebo's approval controls exist for this. Set the ceiling on what an AI employee may do unattended, and route the rest to a person.
4. AI employees acting on your behalf
When you connect an account, the AI employee holds the keys you gave it and nothing more. You are responsible for granting only what the job needs, for what it does with the access, and for revoking it when the job ends. Data that an AI employee obtains through a connected service may be used only to do the work you assigned, and never for advertising, resale, or building profiles of the people it belongs to. Our use of information received from Google APIs adheres to the Google API Services User Data Policy, including the Limited Use requirements, and we hold you to the same standard for the data your employees touch.
An AI employee must not claim to be a human when asked directly, and must identify itself as automated where the law requires it — for example in commercial calls and texts in the United States. It must honor requests to stop, opt out, or be forgotten as promptly as a person would be required to.
5. Content and outputs
AI employees generate text, images, documents, and actions. You are responsible for reviewing outputs before relying on them where accuracy matters, for the intellectual property in what you ask them to produce, and for not passing off generated content as human-authored where that would deceive.
6. Publishers
If you publish skills, plugins, apps, or AI employees to the marketplace, you agree that they will do only what their listing says, request only the permissions they need, handle customer data as this policy and your own privacy terms describe, and stay within the terms of every third-party service they connect to. We review submissions, scan binaries, and remove listings that break these rules.
7. How we enforce this
We do not read your messages or watch your AI employees work as a matter of course. We act on reports, on automated signals such as abuse complaints from providers, spam rates, and unusual traffic, on automated content safety checks, including checks by third-party content safety services, and on what we find when we investigate. A person may review content that is flagged. Flagged content may be kept longer than other data, to protect users and to comply with the law, including reporting child sexual abuse material to the National Center for Missing and Exploited Children. Depending on severity we may warn you, restrict a capability, pause an AI employee, suspend or terminate an account, remove a marketplace listing, or refer the matter to law enforcement. Where practical we tell you what happened and how to appeal by writing to support@neboai.com. We may act without notice when delay would cause harm.
8. Reporting
If you see NeboAI or an AI employee being used against this policy, email support@neboai.com. Include what you saw, where, and when. We read every report.
9. Changes
We will update this policy as the product and the law change. Material changes are announced in the product and take effect when posted; continued use after that is acceptance. The date at the top tells you which version you are reading.
Contact
Questions about this policy: support@neboai.com.