TechnoNetResponsible AI
How TechnoNet builds and uses AI: in our products, in what we claim about them, and in our own engineering.
1. Private by design
EMAS AI platforms are built to run on the customer's own infrastructure, with no third-party AI services in the data path. The customer's documents and decisions stay in the customer's environment.
2. People decide
- Our AI assists people; it does not replace their judgement on decisions that matter.
- In ProjectCore, AI help is optional, can be switched off, and plays no part in staff scoring.
- DecisionCore and AuditCore produce explanations and findings for a responsible person to review. That person decides what to act on.
3. Claims we can show
When we publish an accuracy figure, we publish how it was measured. KnowledgeCore passed 100% (47 of 47 questions) on its published benchmark, with the full method, and 98% (42 of 43) on an academic-policy benchmark. We benchmark accuracy before each delivery.
KnowledgeCore benchmark and method →
We describe regulation as it stands. Malaysia does not yet have a dedicated AI law: the National AI Office's AI Governance and Ethics (AIGE) principles are voluntary, an AI Governance Bill has been proposed and is under consultation, and the PDPA already applies to personal data. We design our products with these in mind and update our wording as the rules change.
4. How we build
TechnoNet engineers use AI development tools, including Anthropic's Claude, to design, write, test and document software. Our engineers review and test every change before release, and TechnoNet remains responsible for everything we deliver.
We use AI development tools on a client's systems or data only as agreed with that client.
Questions
Questions about how we use AI: [email protected]. Personal data: [email protected].