Equitable AI for All

We build AI systems that address bias through diverse data collection and inclusive model design, applying appropriate techniques to reduce unequal impact and ensure fair outcomes for all people and communities affected by our decisions.

Protecting Information and Rights

We process personal, non-personal, and business confidential information in accordance with applicable regulations, implementing strong safeguards to protect against unauthorised access, collection, and misuse, while maintaining confidentiality and protecting individuals’ rights throughout the AI lifecycle.

Understanding How AI Works

We provide clear access to information about our AI system design, functionality, data, and operations, sharing details on algorithms, training data, assumptions, potential biases, risks, and decision-making processes so stakeholders understand how our systems work and which data is used.

Preventing Harmful Decisions

We design systems to manage uncertainty and reduce the risk of harmful outcomes through appropriate safeguards, oversight, and intervention mechanisms when risks arise that may damage brand reputation or negatively impact individuals, society, and the environment.

Resilient Against Threats

We design AI systems to handle unusual conditions, such as abnormal inputs or malicious attacks, without causing unintended harm, defend against vulnerabilities, and ensure reliable performance under difficult conditions through protection against intentional and unintentional interference.

Clear Reasoning for Outputs

We provide clear, understandable reasons for specific AI outputs or decisions, explaining the rationale behind individual outcomes in ways that make sense to people, helping them understand why decisions were made and allowing them to address issues from automated decisions.