Download our thought leadership paper, Trust in new frontiers: Putting AI Governance into Practice and explore how CPAs are steering businesses through AI ambiguity, paving the way for an era of innovation. Join us in reshaping the future of technology governance.
Global public company CEOs and institutional investors are already integrating AI into their operations, with nearly half of them already strategically employing its power.
However, the challenge for the tech industry in explaining AI outcomes has raised concerns, leading to a public confidence crisis. A 2024 Telus survey highlights widespread worry about AI governance among Canadians, with a strong consensus favouring regulation.
As AI models continue to gain adoption, concerns over ethics and transparency are likely to grow louder.
To restore trust and sustain industry growth, forecasted by Bloomberg to reach US$1.3 trillion by 2032, a balanced and pragmatic approach to AI governance is necessary.
“The biggest challenge is the veracity of the data. The pace of change is huge, it’s evolving faster than the assurance models can keep up.”
The Canadian regulatory landscape is constantly evolving. With the release of the Digital Charter Implementation Act (Bill C-27), the federal government is aiming to enhance data privacy, consumer protection, and AI governance.
Proposed amendments to the Artificial Intelligence and Data Act (AIDA) under Bill C-27, include an accountability framework for managing AI systems, risk identification, data protection policies, and employee training protocols.
However, there remains a significant question for policymakers and regulators on how to balance robust governance and controls with flexibility to account for uncertainty, innovation and change in these emerging technologies.
Given the growing importance of AI, future-oriented experts connected to the accounting profession have been calling on CPAs to reposition themselves in the intangible economy and apply their technical skills and Code of Professional Conduct to AI governance.
“AI is not only a technological issue. It is a governance issue for organizations, audit and risk committees and boards.”
Real-world cases, such as Air Canada’s legal battle over AI chatbot errors, underscore the urgent need for organizations to adopt risk mitigation strategies for AI.
A survey published by KPMG in June 2023 reported only an estimated 6% of executives had a dedicated team to evaluate AI risks and implement mitigation strategies, while 47% were still in the initial stages of risk assessment.
As executives emphasize the importance of defining controls over AI, covering aspects like model explainability and fairness to mitigate bias, CPAs can use frameworks like COSO ERM to assess and prioritize AI-related organizational risks.
However, ongoing education is critical for CPAs who are looking to take advantage of the coming market for AI assurance. To compete, CPAs will need greater exposure to both the technical and ethical aspects of AI systems. Without that preparation, they will lose market share to others who are more ready to meet the demand.
“If your job is to oversee the material elements of the organization and this is a material force, then you need to update your skills or get someone else to do it.”
Building trust in AI will require collaboration, and we must dive into the shared responsibility of developing AI models aligned with human values, emphasizing the pivotal role CPAs play in this evolving space.
Beyond the world of tech, many more minds are needed to think through the worst-case scenarios involving AI and develop the measures needed to avert bad outcomes and build confidence in the technology.
Professional accountants globally are adapting their codes of ethics to address the trust gap surrounding AI. As CPAs extend their use of data, discussions are underway about ensuring the growing relevance of ethical codes.
CPAs’ Code of Professional Conduct is a valuable moral compass as AI adoption is happening faster than risk assessment frameworks can evolve. Such a strong ethical code can provide a reliable foundation as teams develop trustworthy AI solutions.
“Responsible AI cannot just fall entirely on the shoulders of data scientists. It requires a combination of different skill sets. The risk and controls mindset, the ethics mindset and the tech mindset.”
CPA Ontario offers a wide range of professional development resources to help CPAs stay up to speed and develop their AI governance capabilities.
Courses on mastering digital technologies and governance.
