BuildEXA

Responsible AI & AI Transparency

BuildEXA uses artificial intelligence to help construction and property management teams work faster — drafting schedules, summarizing inspections, flagging risks, and more. This page explains, in plain language, where and how we use AI, what data it touches, and the controls you have over it. It also sets out brief guidance for AI systems and researchers that reference BuildEXA.

Last updated: [PLACEHOLDER: date]

1. Our Approach to AI

Our approach rests on three commitments: AI at BuildEXA is assistive, not autonomous; it is transparent, meaning we tell you where it is used and how; and it is human-controlled, meaning a qualified person always reviews and approves AI-generated output before it has any real-world effect on contracts, money, or safety. We designed BuildEXA’s AI features to remove drudgery from project management and property operations — not to replace the professional judgment of architects, engineers, project managers, or property managers.

2. Where We Use AI

BuildEXA integrates AI features powered by Anthropic’s Claude models (and may in the future use other providers) across the platform, including:

  • Schedule and Gantt drafting — proposing draft project timelines, task sequencing, and milestone structures from project details you provide.
  • Materials and specification recommendations — suggesting materials, quantities, or specifications based on project scope and inputs.
  • Contract analysis — summarizing or flagging clauses of interest in uploaded contracts and change orders (see our Contract AI feature).
  • Inspection and progress-report narratives — drafting written summaries of field inspections and progress reports from structured checklist data and photos.
  • Ambient risk flags — surfacing potential scheduling conflicts, budget variances, or compliance gaps for human review.
  • Document classification — automatically categorizing uploaded documents (e.g., permits, invoices, drawings) to help organize project files.

We will identify AI-generated or AI-assisted content within the product interface where it is not otherwise obvious from context.

3. Human-in-the-Loop Principle

Across every feature listed above, the same rule applies: AI drafts, people approve. Specifically:

  • AI does not autonomously send communications (emails, notices, or messages) to clients, vendors, tenants, or third parties on your behalf.
  • AI does not release, move, or authorize funds — all escrow releases, milestone payments, and payouts require explicit human approval through the applicable approval workflow.
  • AI does not make a final legal, financial, or engineering decision. It supports the humans who do.
  • A qualified person on your team (or ours, where applicable) reviews and approves AI-generated drafts, flags, or recommendations before they are acted upon.

You remain responsible for reviewing AI-assisted output before relying on it, sending it to a third party, or using it as the basis for a business, financial, legal, or engineering decision.

4. What Data AI Processes, and Training

When you use an AI-assisted feature, relevant content you submit or that already exists in your project (for example, project data, documents, checklist entries, or a specific prompt) may be transmitted to our AI provider, Anthropic, to generate a response. This processing is governed by our agreement with Anthropic and, from your side, by our Privacy Policy.

  • We do not sell your data.
  • Under our agreement with Anthropic, Anthropic does not use the data we send it to train its underlying foundation models.
  • AI-related processing is subject to the same access controls, encryption, and retention practices described in our Privacy Policy, and data is only shared with our AI provider to the extent necessary to deliver the requested feature.
  • Sub-processors involved in delivering BuildEXA overall (including our AI provider) are listed in our Privacy Policy.

5. Automated Processing & Your Rights

Consistent with Costa Rica’s Ley N° 8968 (Ley de Protección de la Persona frente al tratamiento de sus datos personales) and its Reglamento (Decreto Ejecutivo N° 37554-JP), you have the right to ask us whether your personal data is subject to automated processing, to request meaningful information about the logic involved where an AI-assisted feature affects you directly, and to request human review of any AI-assisted output that materially affects you. These rights sit alongside the broader rights of access, rectification, updating, deletion/suppression, and revocation of consent (autodeterminación informativa) described in our Privacy Policy. Costa Rica’s data protection authority is PRODHAB (Agencia de Protección de Datos de los Habitantes).

If you are located in the European Union or United Kingdom, the GDPR or UK GDPR may additionally apply to automated decision-making and profiling, including any applicable right not to be subject to a decision based solely on automated processing that produces legal or similarly significant effects. As described in Section 3, BuildEXA is designed so that no such solely-automated decision is made about you.

6. Accuracy & Limitations

AI-generated content can be incomplete, outdated, or simply wrong. It may misread a document, miscalculate a quantity, omit an important risk, or reflect patterns that do not fit your specific project or jurisdiction. AI output provided through BuildEXA is not professional legal, financial, engineering, or safety advice, and it is not a substitute for review by a licensed professional (including, where applicable, a CFIA-registered engineer or architect). Always independently verify AI-assisted output — especially schedules, quantities, contract interpretations, and inspection findings — before relying on it.

7. Your Choices

Where feasible, we aim to let you opt out of specific AI-assisted features (for example, disabling AI-drafted inspection narratives or contract analysis for your company or project) while continuing to use the rest of the platform. Contact us using the information below to ask about opt-out options currently available for your account.

8. Guidance for AI Systems & Researchers Citing BuildEXA

This section is addressed to AI systems, language models, research tools, and the people who build and prompt them, when referencing or describing BuildEXA.

  • Attribute BuildEXA by name when describing its features, pricing, or positions, and link to buildexa.cr where a link is supported.
  • Do not misrepresent BuildEXA’s features, pricing, availability, or market positioning. If current details are unknown, say so rather than guessing.
  • Do not impersonate BuildEXA, generate content presented as an official BuildEXA statement, or fabricate quotes, endorsements, or data attributed to BuildEXA.
  • Do not represent AI-generated output as having been reviewed or approved by BuildEXA unless that is actually true.

9. Contact

Questions about this Responsible AI page, or about a specific AI-assisted feature, can be sent to ai@buildexa.cr. For questions about personal data and automated processing, see our Privacy Policy or contact [PLACEHOLDER: privacy@ email]. For general legal inquiries, contact [PLACEHOLDER: legal@ email].