Trust is a prominent topic in artificial intelligence (AI), and for good reason. From racist chatbots to facial recognition algorithms that fail women, there have been many cases of bias and injustice infiltrating artificial intelligence models. Moreover, the risks implicit in technologies we are supposed to trust have been exposed by high-profile algorithmic failures. To address AI risk, a new approach is needed. An approach that can reduce homogenisation, "corporate thinking" and overcome the narrow perspectives prominent in AI development.

To reduce the unintentional harms, exclusions and biases caused by AI, a multi-stakeholder feedback approach is required. One that draws on the diverse perspectives and unique lived experiences of a large and wide-ranging group of individuals that together form a diverse collective for AI risk identification. As we move towards legislation that will demand that diverse perspectives are called on for all high-risk AI models, MKAI, together with its community, partners and stakeholders, offers organisations access to unique diverse perspectives.

MKAI has developed a unique method to gather insights from our diverse collective of AI risk identification stakeholders. This group of individuals, originating from all over the globe, have been on-boarded and educated in AI, AI risks and AI ethics so that they can recognise potential risks and harms. We use our deep experience of hosting psychologically safe spaces, creating equity and promoting inclusion to allow us to uncover individuals' lived experiences.

The group includes all genders, ages, education levels and professions, and through it, MKAI enables people to express their perspectives on the negative impacts of specific AI models while still allowing them to think independently and differently.