Does Autoesta write AI policies?
Yes. Autoesta writes an AI policy for each client as part of paid AI consulting, fitted to the brand: its voice, its customers, the data it holds and the AI tools its teams use. We then set those rules inside the agents and workflows we build.
The policy is one output of our AI consulting work, written next to the roadmap that says what to build first. If you want to know what any company AI policy should contain, section by section, our page on whether your company needs an AI policy covers that in full. This page covers what we hand you and how the rules end up inside your tools.
What does Autoesta deliver when it writes your AI policy?
You get a written policy in your company's name, a list of every AI tool your teams use sorted into approved, needs review and stop, the customer-facing wording in your brand's voice, and a list of where each rule is set inside your systems.
- AI tool inventory. Every AI tool in use across your teams, including free ones on personal accounts, sorted into approved, needs review and stop.
- The written policy. Plain language, named after your real tools and your real data, with one owner and a review date.
- Customer-facing wording. The line your chat or voice agent uses to say it is an AI, the line it uses to hand over to a person, and the topics it must never answer, all written the way your brand already speaks.
- Approved answer list. What your agents may say and promise, drawn from approved sources only, with unknown questions sent to a person.
- Compliance notes. Which rules apply to your business and where a person must approve an output. Our pages on US state AI laws and whether the EU AI Act reaches a US company explain the rules we check against.
- A build map. For each rule, the setting in your systems that enforces it, so the policy does not depend on anyone remembering it.
How does Autoesta build the policy into your tools?
Each rule in the policy becomes a setting in the agent or workflow we build on your accounts: a disclosure line, a handover trigger, an approved answer list, a human approval step, logging and error alerts. Then we test those settings with real past examples before customers see them.
A policy that only lives in a document fails the first busy week. This is how the rules map to the build:
| Policy rule | Where we set it | How you can check it |
|---|---|---|
| Customers are told they are talking to an AI | Opening line of each chat and voice agent, in your brand voice, plus a set answer when someone asks | Read a test conversation from the handover pack |
| A person can always take over | Handover triggers and a fallback to a human on every channel | Ask the agent for a person and watch where the conversation lands |
| The AI only promises what you approve | Approved answer list; answers limited to approved sources; unknowns go to a person | Ask it something outside the list |
| High-risk replies need a person | A human approval step before the output is sent or saved | The approval queue in your CRM or inbox |
| Customer data stays where the policy says | The agent reads only the fields it needs; API keys sit in your accounts; the model is chosen on data handling as well as quality, speed and cost | Your own account settings and key list |
| Records are kept | Logging on every run and a run report | The log in your automation tool |
| Failures reach a named person | Error alerts from day one | A test failure during handover |
Every new build uses current models, never one with an announced retirement date, so the policy does not point at a model that is about to disappear. Our page on which AI models Autoesta uses explains how we choose. The disclosure wording itself follows the rules in whether you have to tell customers they are talking to AI. We set these rules inside the AI agents and AI workflows we build, and we test them the way our guide on how to test an AI agent describes.
What does "fitted to your brand" mean?
It means the policy starts from your business instead of a template: how your brand talks to customers, who those customers are, what data you hold, and which AI tools your teams already use. Two companies in the same industry can end up with different rules.
- Voice. A formal firm and a casual one should not use the same disclosure line. We write it in yours.
- Customers. What they are told, when they reach a person, and which topics the AI leaves alone depend on who they are.
- Data. The data you hold sets the strictest rules. Health records, payment details and client files each change what may go into a prompt.
- Tools in use. The policy names your actual tools. A rule written for "AI tools" in general is hard for staff to follow.
How does the work run, step by step?
A free strategy call, a free first audit, then paid consulting quoted for your business. We write the policy with the roadmap, build its rules into your tools on your accounts, test with real examples, and hand over with 6 months of free fixes.
- Free strategy call. Thirty minutes on where AI is used today and where it worries you. See what happens on the call.
- Free first audit. We list the AI tools, data and customer channels in use.
- Paid consulting. Quoted before work starts. The policy is written alongside the roadmap and its compliance notes.
- Review. You read the draft. Any section on hiring, credit, housing or health decisions should also go to your lawyer.
- Build. On your own accounts, with the settings from the build map, at a fixed price shown before work starts.
- Test. Real past examples go through the system, including the awkward ones, before customers do.
- Handover. Written notes on what each part does, plus 6 months of free fixes to anything we built.
How much does it cost to have Autoesta write your AI policy?
At the time of writing, the policy is part of paid AI consulting, which is quoted per business by size and scope. The strategy call and first audit are free. Building the rules into your tools is priced like any build: about $1,000 for a small one.
Full setups run $2,000 to $5,000 and larger systems get a fixed price after the audit, shown before work starts. Optional monthly support from $200 starts after the 6 free months. Our page on how much Autoesta charges for AI consulting breaks the numbers down.
When do you not need Autoesta to write your AI policy?
If you are a small team using one or two AI tools, with no AI talking to customers and no decisions about people, you can write a short policy yourself from a good outline. You also do not need us if your legal team already wrote one and only the tool settings are missing.
- Small, internal AI use only. Use the section outline on our AI policy guide and write it in an afternoon.
- Policy already written. Ask us for a build quote to set the rules inside your agents and workflows, and skip the consulting.
- Legal opinion needed. We are not a law firm. If you need advice on liability or a regulator's view, hire a lawyer. Our page on whether a company is liable for what its chatbot says shows why that matters.
- Company-wide governance programme. A board-level AI strategy across many divisions is often a better match for a large consulting firm.
More questions about Autoesta and AI policies
Can Autoesta write our AI policy without building anything?
Yes. The policy comes with paid AI consulting, and you can stop after the consulting. The written roadmap and policy are yours to keep and use with any provider, including your own team. If you later want the rules set inside your agents and workflows, we quote that build separately, at a fixed price shown before work starts.
Does the policy cover AI tools that Autoesta did not build?
Yes. The inventory covers every AI tool your teams use, including free tools on personal accounts, because those are where client data most often leaks. Each tool goes into approved, needs review or stop. We only set rules inside systems we build or have access to; for other tools, the policy tells staff what is allowed.
Is an AI policy from Autoesta legal advice?
No. We write the policy from the rules that apply to your business and from how your systems work, and our compliance notes point to the official sources. We are not lawyers, so any section on hiring, credit, housing, insurance or health decisions should be read by your own lawyer before you adopt it.
Who owns the AI policy and the systems behind it?
You do. The policy is written in your company's name, and everything we build runs on your accounts, with your API keys and your data. If you change providers, the policy and the settings stay with you. Our page on who owns the automations and data covers the detail.
Does Autoesta write AI policies for large companies too?
Yes. We write AI policies for small firms and large companies, and the method is the same: inventory the tools, write rules that fit the brand, and set them inside the systems. A larger company has more teams, tools and data types in scope, so the consulting quote is larger. See the full list of Autoesta FAQs or read about our team.
