More than one third of Canadian businesses experimenting with ChatGPT, KPMG Canada survey finds

​While more than one-third (37 per cent) of Canadian companies are looking at ways the popular generative artificial intelligence (AI) platform ChatGPT can improve their operations, it pales in comparison to the nearly two-thirds (65 per cent) of U.S. companies using the technology, finds new research by KPMG in Canada.

Overall adoption of AI among Canadian organizations is less than half of that in the U.S., with only 35 per cent of Canadian businesses saying they currently use AI in their operations versus 72 per cent in the U.S. Of the Canadian organizations using AI, more than four in 10 are utilizing it in their call centres, and more than half (53 per cent) of respondents admit they could be using AI more effectively and efficiently.

The findings are based on a survey of senior business and technology leaders at large Canadian and U.S. companies and serves to assess their use of AI, which is increasingly becoming a competitive advantage for businesses. While generative AI is a specific type of artificial intelligence that uses algorithms to produce content quickly, artificial intelligence is much broader in scope, and various forms of AI have been used by Canadian and U.S. organizations in healthcare, education, law enforcement and financial services to boost efficiency and productivity.

"Canadian organizations are trailing behind their U.S. counterparts when it comes to AI adoption, and it comes at a time when developments in the space are moving fast – particularly in generative AI," says Benjie Thomas, Canadian Managing Partner, Advisory Services at KPMG in Canada.

"Generative AI can be powerful if used correctly and responsibly, and it enables businesses to be more efficient, productive and competitive. Canadian businesses that aren't adopting AI today might be putting themselves at risk of falling behind, especially as competitors south of the border continue to advance in this field," adds Mr. Thomas.

The main challenges preventing organizations from using AI effectively are lack of skilled talent and quality data to train AI algorithms. Nearly half (47 per cent) of Canadian businesses said they lack the in-house expertise to validate and verify the accuracy of their AI algorithms, and more than four in 10 (44 per cent) said their data sets are either too small or too big, are missing data, are incorrect or not properly formatted.

"Large organizations are sitting on vast amounts of data, but to make that data work effectively for them, they need to properly assess the quality of their data first. Without that, their journey towards being AI-ready will be rocky, so the first step for any organization thinking about adopting AI is 'get your data ready,'" says Zoe Willis, Partner and National Data and Digital Lead, KPMG in Canada.

Ms. Willis suggests the following steps for organizations to achieve AI readiness:

• Compile a full inventory of the data
• Assess how accurate, relevant and recent the data is
• Identify any outstanding data gaps, and potential/future data gaps
• Evaluate, locate and map out where data sits within the organization
• Determine who has access to the data, and how well-equipped they are to analyze it

"Without quality data, AI algorithms are susceptible to output that is biased, incorrect, misleading and unreliable, and the consequences for businesses include errors that lead to poor business decisions," adds Ms. Willis.

A majority of Canadian businesses are aware of the risks of having poor quality data, with more than half (54 per cent) admitting they are very concerned their organization might be making decisions based on poorly designed AI algorithms, and yet only 44 per cent regularly retaining independent third-party experts to vet or assess their AI algorithms for errors and bias.

"In addition to having quality data, organizations need to adopt strong AI operating models and frameworks that include policies, practices and tools to manage their risk," says Kareem Sadek, Partner, Advisory, IT & Emerging Technology Risk Leader.

"Organizations that want to build their AI capabilities need to have a comprehensive and robust governance plan that addresses data integrity, privacy, reliability, accountability, fairness and security, among other considerations," adds Mr. Sadek, who co-leads KPMG in Canada's Responsible AI Framework.

The survey found only 43 per cent of Canadian organizations have instituted responsible AI frameworks (72 per cent in the U.S.), while half of those believe that their organization strictly adheres to it (70 per cent in the U.S.).

"This highlights a critical need for more organizations to adopt responsible AI frameworks, but also mechanisms to help stick to those frameworks," says Mr. Sadek. He adds that "monitoring, evaluation and continuous evolution are also crucial to any responsible AI framework.

"Artificial intelligence is evolving quickly – as we've seen with ChatGPT, for example – so organizations need to have AI models that are effective, long-lasting but also agile enough to adapt to the world around them. Organizations that don't do this will be less competitive and trustworthy and will eventually fall behind," says Mr. Sadek.

KPMG in Canada researched the opinions of 250 companies in Canada and the U.S. about artificial intelligence between February 21 and 26, 2023, using Sago's Methodify online research platform. Of the 250 companies, 90 are based in Canada and 160 are in the U.S. Fifty-five per cent of all respondents identified as CEOs, business owners, IT Directors and Data Scientists. Thirty per cent reported annual gross revenue of $500 million up to $1 billion, 13 per cent reported revenue of $2 billion up to $5 billion, and 14 per cent reported revenue of $5 billion up to $10 billion.