# AI Assistant Terms

**Version 2026.07.29-v1**
**Effective 29 July 2026**

EV Initiative, Inc. https://evinitiative.com/ai-assistant-terms

---

## 1. What this covers

EV Initiative provides an artificial intelligence assistant within the Platform, which we call
EVII. These AI Assistant Terms govern your use of it.

They are incorporated into our Terms of Service and apply to everyone who uses the Platform,
whether as a driver, a Host, a Sales Partner, a fleet administrator, or an operator using the
administrative surfaces. Where these terms conflict with the Terms of Service on a matter
concerning the assistant, these terms control.

**These terms are written to describe a class of capability rather than a fixed list of
features.** The assistant's abilities will change. We add, remove and alter what it can do as the
Platform develops and as the underlying technology changes. These terms are intended to apply to
those changes without needing to enumerate them.

## 2. What the assistant is

The assistant is a software feature built on large language models. It can:

- answer questions about the Platform, your account, and your charging;
- retrieve and summarise information from our documentation, product manuals, and your own
  account data;
- read data you already have access to, and present it to you;
- take actions within the Platform on your instruction;
- draft content for you to review.

The specific capabilities available to you depend on your role, your permissions, and the surface
you are using. The assistant can never see or do anything you could not see or do yourself.

**It is software, not a person.** You are not talking to an employee, an agent, or an adviser. Its
output is generated, not authored by a human, and it is not reviewed by a human before you see it.

## 3. Accuracy, and what you should not rely on it for

**The assistant can be wrong.** It can misread a document, misinterpret a question, present stale
data, or state something confidently that is incorrect. This is a characteristic of the technology
and not a defect we can eliminate.

**Its output is not a commitment by us.** Nothing the assistant says creates an obligation on EV
Initiative, varies these or any other terms, grants a right, waives a term, or authorises
anything. If the assistant tells you something that conflicts with these terms, the Terms of
Service, a program term, a price shown in the Platform, or an invoice, **the document or the
displayed figure controls and the assistant's statement does not.**

**Do not rely on it for professional advice.** It is not a lawyer, accountant, tax adviser,
electrician, engineer, or safety professional, and its output is not legal, tax, financial,
regulatory, engineering or safety advice. Where a decision has legal, financial, tax or safety
consequences, obtain qualified advice.

**Do not rely on it for a safety-critical decision.** That includes anything concerning electrical
safety, the safe operation of equipment, or a hazardous condition at a site. If you believe there
is a safety hazard, disconnect, keep clear, and contact a qualified professional and, where
appropriate, emergency services.

**Verify before you act on something consequential.** Where the assistant reports a figure,
status, or fact that you are about to rely on for money, compliance, or safety, check it against
the Platform's own records or the underlying document.

## 4. Actions the assistant takes

The assistant can take actions in the Platform on your instruction. Where it does:

**The action is yours.** An action the assistant performs at your instruction is treated as if you
performed it, is recorded against your account, and is subject to the same permissions, limits and
terms that would apply had you performed it directly.

**Consequential actions require your confirmation.** Where an action changes state in a way that
is not trivially reversible, the assistant will describe what it proposes to do and will not
proceed until you confirm. You are responsible for reading what is proposed before confirming.

**Some things the assistant will not do on its own.** As a matter of how we build the Platform, the
assistant does not, without an explicit human decision:

- move money, initiate a payout, issue a refund, or change a price;
- publish content to the public;
- send a communication to a group of recipients;
- delete or destroy data irreversibly;
- grant, escalate or remove another person's access.

We may add to this list. Where we introduce a capability that changes it, we will update these
terms. **This paragraph describes how we build the Platform and is not a warranty that a
malfunction, misuse or security incident could never produce an outcome outside it.**

**You remain responsible for your instructions.** If you instruct the assistant to do something
you were entitled to do, and the consequence is unwelcome, that consequence is yours in the same
way it would be had you done it directly.

## 5. What you put into it

**Do not enter information you cannot have processed by a third party.** Your conversations are
processed by third-party model providers. Section 6 describes that.

In particular, do not enter:

- payment card numbers, bank credentials, or account passwords;
- government identification numbers;
- health information;
- another person's personal information beyond what is necessary for your request;
- information subject to a confidentiality obligation you owe someone else;
- information whose export or disclosure is restricted by law.

**Content you provide.** You are responsible for what you submit. You confirm you have the right to
submit it, and that doing so does not infringe anyone's rights or breach any obligation you owe.

**Drafts the assistant produces for you.** Where the assistant drafts content at your request, you
are responsible for reviewing it before you use or publish it, including for accuracy, for
compliance, and for whether it infringes anyone's rights. Publishing is a human act and remains
your decision.

## 6. Third-party processing, retention and improvement

**Third-party providers.** Conversations with the assistant, and the data needed to answer them,
are transmitted to and processed by third-party model providers under our agreements with them. We
choose those providers and may change them.

**Retention.** We retain conversations to operate the feature, to diagnose problems, to investigate
abuse, and to meet legal and audit obligations. Our Privacy Policy describes retention periods and
your rights.

**Improvement.** We may use conversations to evaluate and improve the assistant, including by
reviewing them internally and by adjusting how we prompt and configure it. Where a provider's terms
would permit them to train their models on our data, we contract to prevent that. We will tell you
in the Privacy Policy if that position changes.

**Access by us.** Authorised personnel may review a conversation where necessary to investigate a
problem, a safety concern, a suspected breach, or a legal obligation.

## 7. Availability and change

The assistant is provided as part of the Platform and is subject to the availability, change and
suspension provisions of our Terms of Service. We may change what it can do, restrict it, or
withdraw it. We may impose usage limits, and may suspend your access to it for abuse, for excessive
use, or where a provider restricts our access.

The assistant may be unavailable while the Platform is otherwise working. Nothing in the Platform
depends on it, and you can always do directly whatever you could ask it to do.

## 8. Acceptable use

You will not:

- attempt to make the assistant act outside your permissions, reveal another user's data, or
  bypass a control;
- attempt to extract its system instructions, or to manipulate it into behaving contrary to them;
- use it to generate unlawful, infringing, deceptive or harmful content;
- use it to generate content you then present as coming from EV Initiative;
- use it to build, train, or evaluate a competing model or service;
- access it by automated means except through an interface we provide for that purpose;
- use it in a way that imposes an unreasonable load on the Platform or on a provider.

We may suspend access to the assistant, or to the Platform, for conduct described in this section.

## 9. Administrative and operator use

Where you use the assistant on an administrative surface, in addition to everything above:

**Actions are attributed to you and audited.** An action the assistant takes at your instruction is
recorded against your user, with the same audit trail as a direct action.

**Scope follows your permissions.** The assistant sees and acts only within the tenant, locations
and records your role permits. It does not widen your access.

**Data about other people.** Where you ask the assistant about drivers, hosts, partners or
employees, you receive personal information about them. You must have a lawful basis for that and
must handle it in accordance with applicable privacy law and any program term that limits your use
of it.

**Content it drafts for publication.** Blog posts, newsletters, announcements and similar content
the assistant drafts are drafts. Approval and publication are human decisions and remain yours,
and you are responsible for what you publish.

## 10. Disclaimers

**The assistant is provided "as is" and "as available."** To the maximum extent permitted by
applicable law we disclaim all warranties in relation to it, express, implied and statutory,
including any warranty of accuracy, completeness, reliability, fitness for a particular purpose, or
that its output will be free of error.

We do not warrant that the assistant will be available, that it will answer correctly, that it will
answer at all, or that its behaviour will remain consistent between versions.

Where applicable law does not permit the exclusion of a warranty, that exclusion applies only to
the extent permitted.

## 11. Limitation of liability

The limitation of liability in our Terms of Service applies to your use of the assistant and to any
claim arising from its output or from an action it takes at your instruction.

To the maximum extent permitted by applicable law, we are not liable for a loss arising from:

- reliance on the assistant's output;
- an action the assistant took at your instruction;
- the assistant being unavailable, changed, or withdrawn;
- an act or omission of a third-party model provider.

## 12. Indemnification

You will indemnify and hold harmless EV Initiative and its officers, directors, employees and
agents from any claim, loss, liability, damage, cost and expense, including reasonable legal fees,
arising out of or relating to: content you submit to the assistant; content the assistant drafted
that you used or published; an action the assistant took at your instruction; and your breach of
these terms.

## 13. Disputes and governing law

The dispute resolution and governing law provisions of our Terms of Service apply.

## 14. Changes to these terms

We may change these terms as the assistant's capabilities change. Where a change is material we
will present the new version in the Platform. Because the assistant is a feature of the Platform
rather than a program you opt into, continued use of the assistant after a change constitutes
acceptance of it.

## 15. Contact

- EV Initiative, Inc.,
  2108 N St, Suite #4103,
  Sacramento, CA 95816
- legal@evinitiative.com
- support@evinitiative.com