Acceptable use policy

This policy applies to everyone who uses the Formsieve service ("Service") of Thinking42, Inc., a Delaware corporation, and to anyone they let use their Formsieve key. If you let a client or colleague use your key, you make sure they follow it too. Words defined in the Terms have the same meaning here.

1. Use the Service for spam filtering on your own forms

The Service is for deciding whether submissions to forms (and, where offered, comments) on websites covered by your plan are spam. Use it through the Formsieve plugin or other clients we publish.

You must not:

1. Use the Service as a general-purpose text classifier or AI model, or send it anything other than real submissions from your sites and the plugin's own tests.
2. Use its results to decide on people's eligibility for jobs, credit, insurance, housing, education, healthcare or public services, or for any decision with legal or similarly significant effects, without meaningful human review.
3. Resell, rent, sublicense or white-label the Service or your key; include it in a hosting plan or platform without a Custom plan; or share your key beyond the sites your plan covers.
4. Put your key in browser code or anywhere the public can read it.

2. Data you must not send

Do not use the Service on forms that collect:

• health information, including information covered by HIPAA;
• payment-card or bank-account details;
• passwords or other credentials;
• government identification numbers or identity documents;
• information from or about children under 13 (or the age of digital consent where you are);
• other special categories of personal data (for example religion, ethnic origin, sexual orientation, political opinions, biometrics), unless those fields are excluded from what the plugin sends.

The plugin excludes fields that look sensitive and redacts phone numbers by default. You remain responsible for your forms.

3. Your duties to the people who use your forms

• Have a lawful basis for screening submissions (for most sites, a legitimate interest in keeping spam out).
• Tell people that an automated AI service screens submissions. The plugin can show such a line under your forms; keep it on or tell people another way.
• Describe the screening in your privacy policy. The plugin gives you text to copy.
• Follow the laws that apply to your site and your visitors.

4. Do not attack, probe or copy the AI model or the Service

You must not:

1. Try to get the model to reveal its instructions, questions or configuration, bypass its safeguards ("jailbreaking"), or make it produce anything other than spam results.
2. Benchmark the model or the Service for publication, or scrape, harvest or bulk-collect results.
3. Use results to train, fine-tune, distil or evaluate an AI model, or to build a product that competes with the Service or with the AI providers we use.
4. Get around rate limits, allowances, site limits, fair-use limits or trial limits (for example with several accounts, rotating sites or disposable e-mail addresses).
5. Send text designed to use unusual amounts of AI processing, such as padding, repeated filler or machine-generated text, to the Service.
6. Use more than one Free plan per person or organisation, or use several accounts, e-mail addresses or keys to get more free AI checks or free sites.
7. Create accounts or trials with automated tools or temporary e-mail addresses.
8. Probe, scan or test the vulnerability of the Service, or interfere with it, except under a written agreement with us. Report security issues to support@formsieve.com.
9. Send malicious code to the Service yourself. (Classifying spam that a third party sent to your form, which may contain malicious links or text, is the Service's purpose and is allowed.)

5. AI provider rules

Our AI providers' acceptable-use rules apply to you as well; the current providers and links to their rules are listed at /subprocessors/. Where those rules are stricter than this policy, they apply.

6. Law and sanctions

Do not use the Service in breach of the law, including sanctions and export laws, or from or for a country, region or person that is blocked under the Terms (§21).

7. What happens if this policy is broken

We may warn you, limit or suspend your key or account, or end the Terms, as described in the Terms (§11 and §19). We may also have to act at the request of an AI provider. Where we can, we tell you first and give you a chance to fix the problem.

8. Contact

Report misuse or ask a question: support@formsieve.com.