At Ukweli AI, we believe that artificial intelligence must be developed and deployed responsibly, transparently, and in a manner that respects human dignity and promotes equity. This AI Ethics Policy sets out the principles that guide how we design, build, and use AI systems, and how we hold ourselves accountable to the communities we serve.

1. Human-Centred Design

Every AI system we build is designed with people at the centre. We prioritise user well-being, autonomy, and informed consent. Our technology serves people; people do not serve our technology. We actively involve end-users — including marginalised and underrepresented communities — in the design and evaluation of our systems.

2. Fairness and Non-Discrimination

We are committed to building AI systems that are fair and free from unjustified bias. To achieve this we:

• Rigorously audit training data and model outputs for demographic bias.
• Test systems across diverse population groups before deployment.
• Provide clear mechanisms for affected individuals to report discriminatory outcomes.
• Regularly review and update models to mitigate newly discovered biases.

3. Transparency and Explainability

We believe people have the right to understand how AI decisions that affect them are made. We therefore:

• Clearly communicate when users are interacting with an AI system.
• Provide meaningful explanations of AI-generated outputs in plain language.
• Publish documentation on the data sources and methodologies used to train our models.
• Disclose known limitations and uncertainty in AI outputs.

4. Privacy and Data Protection

Privacy is a fundamental right. Our AI development practices reflect this by:

• Collecting only the minimum data necessary for a given task (data minimisation).
• Applying privacy-by-design principles throughout the AI lifecycle.
• Anonymising or pseudonymising personal data wherever feasible.
• Obtaining explicit, informed consent before using personal data in training datasets.
• Complying with applicable data protection laws, including GDPR and any local regulations.

5. Safety and Reliability

We take the safety of our AI systems seriously. Our commitments include:

• Conducting thorough testing and red-teaming before any public deployment.
• Implementing robust monitoring and alerting for unexpected or harmful outputs.
• Maintaining human oversight and the ability to intervene or shut down systems if needed.
• Never deploying AI in contexts where failure could cause physical harm without extensive safeguards.

6. Accountability and Governance

Responsibility for AI outcomes rests with us, not with the technology itself. To ensure accountability:

• We maintain a dedicated AI Ethics review process for all new AI features and products.
• We designate responsible individuals for the ethical oversight of each AI system.
• We provide accessible channels for users to raise ethical concerns or report harms.
• We track, document, and publicly report on ethical incidents and the steps taken to address them.

7. Inclusivity and Accessibility

AI should benefit everyone. We strive to build systems that are accessible to people with disabilities, work well across different languages and cultural contexts, and do not require expensive hardware or high-speed internet. Reducing the digital divide is core to our mission.

8. Environmental Responsibility

Training and running large AI models consumes significant energy. We commit to measuring and minimising the environmental footprint of our AI workloads by selecting energy-efficient infrastructure, preferring renewable energy sources where available, and reporting on our carbon emissions annually.

9. Prohibited Uses

We will not develop or knowingly allow our AI systems to be used for:

• Mass surveillance, unlawful tracking, or profiling of individuals.
• Generating or distributing disinformation, deepfakes, or manipulative content.
• Autonomous lethal weapons or systems intended to cause physical harm.
• Discrimination based on race, gender, religion, disability, sexual orientation, or any other protected characteristic.
• Circumventing legal protections or human rights obligations.

10. Continuous Improvement

The field of AI ethics is rapidly evolving and so are we. We review and update this policy at least annually, engage with external ethics researchers and community stakeholders, and incorporate new best practices as they emerge. We welcome feedback from anyone who interacts with our systems.

Contact Us

If you have questions or concerns about this AI Ethics Policy, or wish to report an ethical issue, please reach out to us through our Contact page. We are committed to responding promptly and transparently.