Royal Bank of Canada's artificial intelligence research unit has launched a programme to help promote "ethical AI".
While most Canadian businesses think that it is important to implement AI in an ethical and responsible way, the vast majority say they face barriers - such as cost, time and lack of understanding - to doing so.
RBC's Borealis AI unit is hoping to tackle this with its Respect AI online hub, which brings together open source research code, tutorials, academic research and lectures that firms can use.
The vast majority of Canadian companies surveyed by RBC believe they have bias within their organisation, but almost half do not understand the challenges that bias presents in AI.
Just 53% of respondents have someone responsible for ethical development of data and AI technology and only 23% are able to fully explain the decisions and actions taken by their artificial intelligence models.
The new hub will focus on five areas: robustness against adversarial attacks; data privacy; fairness; model governance; and explainability.
Foteini Agrafioti, head, Borealis AI, says: "Responsible and safe AI is critical to maintaining trust and accountability, but as this new survey shows, many companies and developers do not have the resources to implement AI safely and ethically.
"Respect AI will help enable secure, fair, ethical and trusted AI products and a more responsible adoption of AI technology across industries."
Last year RBC joined forces with Western University on a programme exploring the ethical and social aspects of data analysis and AI.