What is ProjectVIEW AI data security and governance? DANAOS ProjectVIEW AI operates inside ProjectVIEW ERP’s governance, not around it. Every AI request follows the chain Identity → Role → Permission → Data → AI → Action, consequential actions always need human approval, and AI activity is audited. Data is not sent to public consumer AI services: ProjectVIEW AI runs in a private environment, and DANAOS Projects holds ISO 27001, ISO 9001, ISO 14001 and ISO 20000-1. This page explains the AI data governance model for ERP data, and where to find the supporting security documentation. What is the governance chain? AI should operate inside enterprise governance, not around it. In ProjectVIEW AI, each step depends on the one before it: Identity. Who is making the request. Role. What that person does in the organisation and on the project. Permission. What the role is allowed to see and do in ProjectVIEW ERP. Data. Only the data those permissions allow is retrieved. AI. The model works on that data and nothing else. Action. Any resulting action goes through ERP validation and, where it matters, human approval. A cost controller and a site engineer asking the same question can therefore get different answers, because they have different access. When does AI need human approval? Consequential actions always need human approval. ProjectVIEW AI agents follow a fixed cycle: Detect → Investigate → Explain → Recommend → Request Approval → Execute → Audit. Example: a procurement agent flags a quotation above the budget rate. It explains the gap, recommends an action, and requests approval from the responsible user. Only after approval is anything executed in the ERP. The same applies to voice and headless capture from site. Voice recordings alone do not establish contractual entitlement or certified progress. They are interpreted, mapped to BoQ, WBS and Cost Codes, and entered subject to validation and authorisation. How are AI actions audited? Audit is the last step of every agent cycle. The record shows what the agent detected, what it recommended, who approved, and what was executed. Managers can review AI-supported decisions the same way they review any other ERP transaction. This matters for claims, disputes and internal control. An AI recommendation that cannot be traced is a liability; one that is tied to source data and an approver is evidence. Where is our data processed and stored? ProjectVIEW AI runs in a private environment. Data is not sent to public consumer AI services. Control What it means Private environment AI runs in a dedicated setup, not in public consumer AI tools. No training on prompts (Azure) Prompts are not used for model training. Data residency options (Azure) Choice of where data is held. Encryption at rest (Azure) Stored data is encrypted. Customer-managed keys (Azure) The client can control the encryption keys. On-premises option ProjectVIEW can run in the client’s own infrastructure. Sovereign KSA (planned) Client-specific environments on the Azure Saudi Arabia region, announced for November 2026. Models retrieve ERP context at inference time, so no retraining of models on your data is required. Deployment choices are covered on ProjectVIEW AI deployment. Our information security approach is described under information security and confidentiality and data protection and privacy. Which certifications does DANAOS Projects hold? ISO 27001: information security management. ISO 9001: quality management. ISO 14001: environmental management. ISO 20000-1: IT service management. Certificates are listed on the ISO certifications page. Who is responsible for decisions made with AI? People are. ProjectVIEW AI detects, explains and recommends. The user with the right role approves or rejects. ERP records what happened; AI explains why it happened and what may happen next; a named person decides what to do about it. This keeps accountability where it already sits in your organisation: with the project manager, the cost controller, the procurement officer and the directors who sign off. What should IT and security teams review? Before rollout, we suggest your IT and security teams review these points with us: Deployment option: Azure cloud, on-premises, or sovereign KSA (planned). Data residency and encryption settings, including customer-managed keys on Azure. Role and permission design in ProjectVIEW ERP, since AI access follows it. Which actions need human approval and who approves them. Whether staff will use the MCP connector with Claude or ChatGPT, and under which business or enterprise workspace. Audit review: who checks AI-supported decisions and how often. The IT manager role page covers the wider ProjectVIEW picture for IT teams. How does governance reduce the risk of AI errors? AI models can be wrong. Governance limits the impact. Because AI only reads permitted ERP data, explains its reasoning and cannot execute consequential actions without approval, an error becomes a recommendation someone rejects, not a posted transaction. AI will not fix broken processes. It will amplify them. That is why ProjectVIEW AI sits on top of an ERP with defined workflows and approval rules. More on this in AI risk management. Frequently asked questions Can ProjectVIEW AI see data a user is not allowed to see? No. The AI works within ProjectVIEW roles and permissions. It retrieves only the data the requesting user is allowed to access. Can an AI agent post transactions on its own? No. Consequential actions always need human approval. Agents detect, investigate, explain and recommend, then request approval before anything is executed. Is our data sent to ChatGPT or other public AI services? Data is not sent to public consumer AI services. If you choose to use the ProjectVIEW MCP connector with Claude or ChatGPT, use your organisation’s business or enterprise workspace under your own agreement with that provider. Can we keep data in a specific country? On Microsoft Azure there are data residency options, and ProjectVIEW can also run on-premises. Sovereign client-specific environments in KSA on the Azure Saudi Arabia region are planned. Which ISO certifications does DANAOS Projects hold? DANAOS Projects holds ISO 27001, ISO 9001, ISO 14001 and ISO 20000-1. The certificates are listed on our ISO certifications page. Need to take this to your IT and security team? Book a demo and bring your IT and security leads. We will walk through the governance chain, approvals, audit and deployment options. Book a Demo