1. Our commitment
At Lido Technologies Inc., we believe AI should be built in a way that earns and deserves trust. As a developer of custom AI and machine learning systems, we take responsibility for how our technology is designed, tested and deployed. The principles below guide every project we undertake.
2. Human oversight and accountability
AI should support people, not replace their judgment on decisions that matter. We design systems so that humans remain in control of consequential decisions, with clear processes for review, override and escalation. Every project has named people accountable for its outcomes, on our side and on our client's.
3. Transparency and explainability
We document how our models are built, what data they were trained on, how they were evaluated and where their limitations lie. Where it matters for a use case, we build in explainability so that users can understand why a system produced a given output. We are honest with clients about what AI can and cannot do.
4. Fairness and bias mitigation
Models can reproduce or amplify biases in the data they learn from. We assess training data for representativeness, test models for unequal performance across relevant groups, and work with clients to mitigate unfair outcomes before deployment, and to keep monitoring for them afterward.
5. Privacy and data protection
We follow privacy-by-design principles and comply with applicable privacy law, including Canada's PIPEDA. We collect and use only the data a project needs, apply de-identification where practical, and protect data with strong security controls. Client data is never used to train models for other clients or for our own products without explicit written consent.
6. Safety, security and reliability
We test our systems for accuracy, robustness and security before release, including, for generative AI, testing against misuse such as prompt injection and harmful outputs. Production systems are monitored for performance degradation and data drift, with alerting and rollback plans in place.
7. Responsible generative AI
When we build applications using large language models, we ground responses in trusted sources wherever possible, apply content safeguards appropriate to the use case, respect user permissions and data boundaries, and recommend clear disclosure to end users when they are interacting with an AI system.
8. Legal and regulatory compliance
The regulation of AI is evolving rapidly in Canada and around the world. We monitor developments in AI governance, privacy and sector-specific rules, and help clients design systems that meet current requirements and are prepared for emerging ones, including international frameworks such as the EU AI Act where they apply.
9. What we won't build
We decline projects intended to deceive or manipulate people, to conduct unlawful surveillance, to discriminate unlawfully, or to cause harm. If concerns arise during a project, we raise them openly with our client and will pause or stop work where necessary.
10. Questions and concerns
If you have questions about our Responsible AI practices, or concerns about a system we have built, please contact us at support@technologies-lido.com or +1 (437) 524-1626. We take every concern seriously.