By business type

Acceptable Use Policy Generator for startups

Written for the stage you are actually at - and for the diligence that will read it later.

Generate your acceptable use policy Read the acceptable use policy guide

An early-stage product does not need a long acceptable use policy, but it does need one before the first abusive account - because suspending a paying customer without a rule to point at is a contract problem as well as a support problem.

Startups get privacy documentation wrong in two opposite directions. Some publish nothing until a customer asks, which means the first enterprise deal stalls in procurement. Others copy a large company’s policy, which describes a data protection officer they do not have, certifications they have not obtained and processes they do not run - commitments that become liabilities the moment anyone checks.

The right document at seed stage is short, accurate and honest about scale. It names the handful of tools actually in use, states real retention periods, and does not claim an accreditation nobody has been through. That document survives diligence; an aspirational one does not.

Diligence is the reason to get this right early. Data room requests routinely include the privacy policy, the processor agreements, the sub-processor list, the breach log and evidence of marketing consent. Reconstructing those retrospectively is expensive, and gaps become price adjustments or indemnities.

What a acceptable use policy for a startup has to cover

Prohibited content and conduct, kept short and genuinely enforceable
Technical limits: rate limiting, scraping and resource abuse
A responsible disclosure route for security researchers
The enforcement steps you will actually take, from warning to termination
How a suspended user gets their data out

How a startup actually moves personal data

Waitlists and landing page sign-ups

Collected long before the product exists, often with no stated purpose beyond "we will let you know", and then used for launch marketing.

Beta and design partner data

Real customer data in an environment with fewer controls than production, frequently under an informal arrangement rather than a contract.

Founder-led sales outreach

Scraped or purchased contact lists, which need a legitimate interests assessment and, under Article 14, notification to people whose data you did not collect from them.

Product analytics from day one

PostHog, Mixpanel or similar instrumented before anyone considered the privacy notice.

Shared credentials and personal accounts

Early-stage teams routinely hold customer data in personal Drive folders and shared logins, which is a security disclosure problem as much as an access-control one.

Investor and diligence data rooms

Customer lists and metrics shared during fundraising, which is a disclosure to third parties needing its own basis.

Third parties the draft will ask you about

Vercel or Render · Supabase or Neon · Stripe · PostHog or Mixpanel · Resend or Postmark · Slack · Notion · Google Workspace

The rules that apply

Transparency from the first user

The obligation attaches to the first person whose data you process, not to a revenue threshold.

Article 28 agreements with your vendors

Every tool holding personal data needs processor terms in place. Most SaaS vendors publish a DPA you can accept without negotiation.

Marketing consent records

Cold outreach and waitlist emails both need a documented basis, and the record is what diligence asks for.

Breach obligations regardless of size

The 72-hour notification clock in the UK and EU has no small-company exemption.

Founder-collected data

Spreadsheets of leads, investor contacts and beta users are processing like any other, with the same obligations.

What the generated acceptable use policy contains

Prohibited content and conduct
The categories you will not host, written specifically enough to enforce and narrowly enough to defend.
Technical restrictions
Rate limits, scraping, load testing, resource abuse and circumvention of quotas.
Security research boundaries
What testing is permitted, how to report a finding, and what will not be treated as an attack.
Reporting and moderation
How abuse is reported, what happens next, and the timescales you work to.
Enforcement ladder
Warning, throttle, suspension, termination - with the emergency route for genuinely urgent harm.
Appeals
How someone contests a decision, which several online-safety regimes now require rather than merely encourage.

The minimum viable compliance set

Publish an accurate, short privacy policy
Naming the tools you actually use and the retention you actually apply.
Accept your vendors’ DPAs
Keep the signed or accepted copies in one folder for diligence.
Start a sub-processor list now
It is trivial with six vendors and painful with sixty.
Record marketing consent from the first email
Source, timestamp and the wording they agreed to.
Write a one-page breach procedure
Who is called, who decides, and the 72-hour clock.
Move customer data out of personal accounts
Before you write a security section that claims access control.

Where this usually goes wrong

Claiming certifications you do not have

A policy promising ISO 27001 or SOC 2 before the audit is a misrepresentation that diligence will find.

Naming a DPO you have not appointed

Copied templates do this constantly. If you do not need one, say who is accountable instead.

Waitlist data used for something it was not collected for

Sign-ups for launch notification are not sign-ups for a newsletter or a sales sequence.

No vendor DPAs in place

Most vendors publish one. Accepting them takes an afternoon and is a standard diligence request.

Customer data in personal accounts

It undermines every security statement in the policy, and it is the first thing a technical reviewer probes.

Cold outreach with no Article 14 notice

Where you did not collect the data from the person, you owe them information about the processing.

Frequently asked questions

Do I need a privacy policy before launch?

If you are collecting waitlist emails, yes - that is already processing. The policy can be short, but it needs to exist and be accurate about what happens to those addresses.

Do I need a DPO as a startup?

Almost certainly not. The threshold is large-scale regular monitoring or large-scale special-category processing. What you should do is name an accountable person instead of copying a DPO clause you cannot honour.

What will investors ask for?

Typically the privacy policy, terms, vendor DPAs, sub-processor list, security summary, breach log and evidence of marketing consent. Having them assembled shortens diligence noticeably.

Is cold outreach legal?

B2B cold email is permitted in more places than B2C, but it still needs a lawful basis, an opt-out, and in the UK and EU an Article 14 notice to people whose data you obtained elsewhere.

Do I need an acceptable use policy separate from my terms?

A separate AUP is easier to enforce and easier to update. It also gives moderation staff and automated systems a single reference to cite, which matters when a suspension is challenged.

Does an AUP help with platform liability?

It is part of the picture. Intermediary liability protections generally depend on acting on notice, and both the EU Digital Services Act and the UK Online Safety Act expect published rules, a reporting route and an appeals process.

How specific should prohibited-use lists be?

Specific enough that a moderator can apply it consistently, with a residual catch-all. Lists that are only catch-alls get challenged; lists that are only specific leave gaps.