Artificial intelligence is a core capability within the Deep Space AI platform.
Our AI system, known as KAI, is designed to assist construction teams byanalysing project information, summarising documents, drafting communications,and providing workflow insights.
Deep Space deploys AI using a strict governance framework that prioritiseshuman oversight, data protection, transparency, and operational safety.
This page explains how AI operates within the Deep Space platform and how customer data is protected.
1. AI as a Copilot, Not a Decision Maker
KAI operates as an AI copilot embedded within projectworkflows across Deep Space modules including Preconstruction, Commercial, Procurement, Delivery, Documents, and HSEQ.
KAI assists teams by:
• analysing project data
• summarising documentation and correspondence
• drafting RFIs, reports, and communications
• identifying insights across project records
AI systems do not autonomously execute actions inside theplatform.
Actions such as issuing RFIs, approving invoices, sending correspondence, creating variations, or distributing documents always requireuser review and confirmation.
2. Human Oversight
Deep Space AI is designed with a human‑in‑the‑loop architecture.
All AI outputs must be reviewed and approved by users beforeany operational action occurs.
Project managers, contract administrators, site supervisors, and other users remain responsible for decisions made within the platform.
AI assists with information analysis but does not replace professional judgement.
3. AI Data Protection and Privacy
Deep Space follows strict data protection principles.
Customer data submitted to AI features is processed solely to generate the requested output.
Deep Space does not use customer data to train public or shared AI models.
Customer prompts, documents, and project data are not pooled across tenants for model training.
Each customer workspace operates within an isolated environment.
4. AI Security Architecture
AI services operate within the same secure infrastructure as the broader Deep Space platform.
Security controls include:
• encryption in transit and at rest
• role‑based access controls
• infrastructure monitoring and logging
• secure cloud architecture
• strict authentication and permissions management
These controls ensure customer data remains protected throughout AI processing workflows.
5. Third‑Party AI Infrastructure
Deep Space may utilise trusted third‑party AI infrastructure providers to generate AI outputs.
Where third‑party providers are used:
• prompts are processed only to generate outputs
• providers are contractually restricted from using customer data for model training
• data protection obligations apply under strict contractual safeguards
6. AI Transparency
Users are informed when AI features generate content or recommendations.
AI outputs are clearly distinguishable from user‑generated content within the platform.
Users maintain full control over whether AI suggestions are used.
7. AI Limitations
AI systems rely on probabilistic models and may produce inaccurate or incomplete outputs.
AI‑generated suggestions should always be reviewed before being relied upon in contractual, financial, engineering, or safety decisions.
Deep Space AI is intended to assist construction teams, not replace professional judgement.
8. Responsible Use
Customers must use AI features responsibly.
Users must not use AI functionality to:
• generate unlawful or harmful content
• attempt to reverse engineer AI models
• extract model training data
• create competing AI systems
Misuse of AI features may result in suspension of access.
9. Continuous Improvement
Deep Space continuously improves AI capabilities through monitoring, testing, and customer feedback.
This includes:
• refining AI models
• improving safety controls
• enhancing workflow integration
• strengthening governance and compliance frameworks
AI technology evolves rapidly and Deep Space regularly updates its AI systems to improve reliability and usefulness.
‍
.png)




.png)





