Our Commitment to Ethical AI
At Optimus, we believe that AI should be a force for good. Our ethical framework ensures that every solution we deliver is responsible, transparent, and designed with human values at its core.
Our Guiding Principles
These six principles inform every decision we make and every solution we build.
Transparency
We believe in being open about how AI systems work. Every solution we implement comes with clear documentation and explanations of its decision-making processes.
Human-Centred Design
AI should augment human capabilities, not replace human judgment. We design systems that keep people at the centre of decision-making.
Inclusivity
Our AI solutions are designed to be fair and accessible, avoiding bias and ensuring equitable outcomes for all users and stakeholders.
Accountability
We take responsibility for the AI systems we help implement. Clear ownership and governance structures ensure ongoing oversight.
Privacy & Security
Data protection is paramount. We implement robust security measures and ensure compliance with privacy regulations in every project.
Sustainability
We consider the environmental and social impact of AI implementations, promoting solutions that are sustainable long-term.
Our Promise
We're committed to staying at the forefront of ethical AI practices. This means continuous learning, regular audits of our recommendations, and honest conversations with our clients about the implications of AI adoption.
If you have questions about our ethical approach or want to discuss how we ensure responsible AI in your project, we'd love to hear from you.
Frequently Asked Questions
- What is your position on AI ethics, in plain terms?
- A person stays accountable for anything that affects a customer, and any decision the system makes has to be explainable afterwards. Those two rules cover most of it. The rest is applying them honestly when it is inconvenient.
- What does "human in the loop" actually mean day to day?
- It means AI drafts and a named person approves before anything reaches a customer or a court. Not a rubber stamp, an actual check by someone who can override it. Where the stakes are low and the volume is high you can relax it, but you decide that deliberately and write it down.
- How do you handle customer data?
- Data minimisation first, so a process only sees what it needs to do the job. No personal or client-confidential data goes into general-purpose public tools. Retention and deletion get specified when the thing is designed, not retrofitted after a data request lands.
- Will you tell us if AI is the wrong answer?
- Yes, and it happens regularly. I do not resell software, so there is no commission riding on the recommendation. If the process needs fixing first, or the problem is a people problem, that is what the report will say.
- What about bias in AI decisions?
- Any system that touches pricing, eligibility or customer outcomes needs testing for unequal treatment before it goes live, and monitoring afterwards. In insurance and claims this is not a theoretical concern, it is a regulatory one. If a client does not want to fund that testing, I would rather not build the system.
- How does the EU AI Act affect a UK business?
- It affects you if you serve EU customers or your systems touch EU data, and plenty of UK businesses do without having thought about it. The practical step is knowing which of your uses are high-risk under the definitions, which is usually a shorter list than people fear.
- Do you disclose when AI has been used?
- Yes, and I would advise you to. Disclosure is cheap and the trust cost of being found out is not recoverable.