AI Product Terms

The terms for the AI features, and how they handle data.

Last Updated: 3 September 2026

About Teralo AI

Teralo uses Google Gemini to deliver Large Language Model capabilities.

AI prompts and responses are sent to the Google Gemini Developer API, which under its paid-tier terms does not use them to train or improve Google products, and are not sent to Teralo's product analytics provider.

Neither Teralo nor Google uses your Input or Output to train Large Language Models. Both hold copies of it: Teralo stores your conversation in its own database in Sydney, under the same access controls as the rest of your project data, and Google logs Input and Output for a limited period in order to detect abuse of its API. Section 5.3 sets this out in full.

1. Introduction

These AI Product Terms apply to your use of AI-powered features through Teralo's platform. Use of AI Products is optional and at your sole discretion.

These terms clarify:

  • Use of AI Products is optional and at your risk
  • Teralo disclaims all warranties for AI-generated content
  • You must ensure compliance with applicable laws
  • Teralo makes no commitments or guarantees regarding AI Products

2. Defined Terms

Input: Any data, information, content, or material you submit to the AI Products

Output: Content or information generated by the AI Products in response to your Input

3. Ownership, License and Use

3.1 Ownership

You retain all rights and intellectual property in your Input and resulting Output.

3.2 License to Teralo

By using AI Products, you grant Teralo a non-exclusive, worldwide, royalty-free license to access, use, and process your Input and Output solely to provide services, including security, maintenance, and improvements.

This licence does not permit Teralo or its AI Technology Partners to use your Input or Output for training Large Language Models.

3.3 Sharing with Third Parties

Teralo may share Input and Output with service providers and authorized third parties as necessary to provide AI Products or comply with law, subject to confidentiality obligations. The providers who receive them are named in section 5.2.

3.4 Your Responsibility

You are solely responsible for the legality and appropriateness of your Input. Do not submit sensitive personal information including racial/ethnic origin, political opinions, religious beliefs, health data, or sexual orientation data.

This restriction is about what you type or paste into an AI prompt. It is not a restriction on what Teralo itself is built to hold. Teralo collects biometric templates for optional kiosk face sign-in, with the worker's express consent, and records incidents and injuries because a construction platform has to. Those collections are governed by the Privacy Policy, not by this section.

4. Restrictions on Use

4.1 Prohibited Uses

You must not use AI Products to:

  • Submit Input or Output infringing intellectual property or privacy rights
  • Include sensitive personal information
  • Create misleading or deceptive content
  • Violate applicable law
  • Generate professional advice (legal, medical, financial)
  • Perform fully automated decision-making without human oversight
  • Create false, defamatory, harassing, or harmful content

4.2 Disclosure Requirement

If you share Output publicly, clearly disclose that content is AI-generated in an easily visible manner.

4.3 Content Sharing and Publication

When publishing Output, you must:

  • Manually review and verify each Output
  • Attribute content to your name or entity
  • Disclose AI-generated content
  • Not use Output to impersonate or mislead

5. Output Limits and Technology Partner Disclosure

5.1 Output Volume Limits

Teralo may impose reasonable limits on volume, frequency, or nature of generated Output based on your subscription plan.

5.2 AI Technology Partners

Two third parties are involved in an AI call, and they receive different things:

Google, through the Gemini Developer API, generates the Output from your Input. It is the only third party that receives either.

PostHog, Teralo's product analytics provider, records that a call was made and how it performed: the model, token counts, cost, latency, and whether it succeeded. It does not receive your Input or your Output, and it does not receive the records the assistant read in order to produce the Output. Teralo uses those measurements to monitor the health of AI features. PostHog processes them in the United States.

Both are listed, with the categories of data they receive, on the sub-processor page.

5.3 Data Handling and Privacy

Training. Neither Teralo nor Google uses your Input or Output to train Large Language Models. Google's terms for the paid tier of the Gemini API state that Google does not use prompts or responses to improve its products.

Retention by Google. Google logs Input and Output for a limited period for the sole purpose of detecting and preventing abuse of its API. Google does not publish a fixed number of days for that period, and Teralo does not represent one.

Retention by Teralo. Teralo retains your conversation in its own database until you delete it or the project is deleted. There is no second copy held with an analytics provider, so deleting a conversation ends Teralo's retention of it. A stored conversation is held under the same access controls, confidentiality obligations and security measures as the rest of your project data, which are set out in the Data Processing Agreement.

Processing location. Teralo uses the Gemini Developer API, which does not offer a regional processing commitment. Your Input may be processed on Google infrastructure outside your own country, including outside Australia. Teralo makes no representation that AI processing is performed in any particular region.

6. No Professional Advice Disclaimer, Authority Disclosure, and Feedback

6.1 No Professional Advice

Output is generated by AI and provided for informational purposes only. Teralo makes no representations regarding accuracy, completeness, or reliability. Output does not constitute legal, financial, medical, or professional advice.

6.2 AI Limitations

AI systems may struggle to understand context, lack human judgment, perpetuate biases, generate inaccurate content, and face challenges with complex reasoning. You must evaluate and verify all Output before relying on it.

6.3 Prohibition on Fully Automated Decision-Making

Do not use AI Products for decisions with legal or significant effects on individuals without appropriate human oversight, including employment, credit, insurance, housing, or legal matters.

6.4 Disclosure to Authorities

Teralo may disclose Input and Output to law enforcement or authorities if reasonably necessary to comply with law, respond to claims, investigate unlawful use, or protect rights and safety.

6.5 Feedback Mechanism

You may submit feedback on AI Products. Feedback may be used by Teralo to improve products and is not treated as confidential.

6.6 Disclaimer

To the maximum extent permitted by Law, Teralo excludes all liability for any Loss, damage, claim, cost, or expense arising out of or in connection with your use of the AI Products or any Output.

You use AI Products at your own risk. Teralo does not warrant accuracy or fitness for purpose. You are solely responsible for verifying Output suitability.

Contact

For questions about these AI product terms, please contact us at support@teralo.co.

Every version of the AI Product Terms stays published at its own address, unchanged. An agreement that names a version by its date is governed by that version's text.