CPA Ontario recently issued Trust in New Frontiers: Putting AI Governance into Practice.

The foreword states: “The question of trust is at the heart of the debate surrounding artificial intelligence and its impact. And as the use of AI becomes even more commonplace in business and society, that question of trust is also at the core of how CPAs will respond to the challenge of AI… To realize the true potential of AI, we must put integrity, the protection of the public and building of trust at the top of our priority list.” The paper lays out “how CPAs have the skills and the ethical mindset needed to build trust in AI.” It concludes:

• So this is a call to action for all CPAs to join us in charting this new frontier. To pursue the skills, tools and education you need to play this important role in a rapidly changing profession and a rapidly changing world. Trust is the bedrock of our profession, and it is the key to unlocking the economic potential of AI.
• For the record, this paper was researched, written and reviewed by humans. But one day, very soon, the question of whether generative AI was used in the writing of this paper will no longer matter.

That much-repeated term “trust” is almost a mandated component of any CPA “thought leadership” publication. For example, a CPA Canada publication of a few years ago on distributed ledger and blockchain technologies used the term several dozen times. For one, CPAB’s Carol Paradine wrote that whereas auditors inject trust into the traditional financial reporting process “by attesting that the financial statements are fairly stated…we might be tempted to think that, with trust inherent in the programming of the smart contracts and the immutability of the blockchain record, there is not much left for an auditor to do. (But) I don’t think that we are there yet. I strongly believe that a (human) auditor’s judgment and professional skepticism will not be replaced by machines any time soon. We are somewhere in between, with auditors collaborating with ever-smarter machines.”

Likewise, a more recent CPA Canada publication on sustainability noted: “The assurance profession will need to play a significant role in enhancing trust and confidence in the sustainability information that entities are disclosing and in identifying opportunities for improvement.” And a CPA Canada article on cryptocurrency noted several times how accountants might act as “trusted advisers” in the cybersecurity area. More broadly, a Pivot magazine article concluded with a thought on CPAs “applying their core competencies of trust and ethics to the Big Data world at home and abroad. “

There’s something recurringly strange though about the CPA profession perpetually telling its members, in effect, to tune into yet another strange and evolving area, because if they knew anything about it, their involvement could make it more trustworthy for others. Even for the best-established mechanism of accountant-injected trust, the annual audit, it’s hard to truly gauge and value the additional trust and confidence provided by the audit report, because we don’t have access to a parallel control-group universe in which external assurance doesn’t exist, or does so only in a lesser form.

The Reagan-era maxim “Trust, but verify,” could well be positioned as a reference point for discussions about accounting and assurance. It contains something of a paradox though: at a certain degree of verification, it may be largely irrelevant (or at least not determinative) whether one trusts or not. Something approaching half the US electorate still apparently places more trust in Trump than in his opponent, despite rampant evidence of unsuitability (for those inclined to believe his ridiculous assertion that the crowd photos of Kamala Harris’ rallies were AI-generated, for example, no amount of CPA assurance would ever have convinced them otherwise); stock prices are as likely to respond to impression and rumour as to verifiable information. For new waves of technology, the prevailing cutting-edge attitude is generally along the lines of adopt and appropriate and exploit now and fill in the compliance details later; consequently, the proliferating adoption of AI goes hand in hand with the pervasive (often even existential) fear of it. Overall, human lives have become so fraught and complicated and risk-laden and plain difficult that the question of whether to trust any particular aspect of our experience may seem almost quaint.

We see in many contexts how the chase for financial returns in a challenged world outpaces any considerations of prudence and sound valuation, let alone of ethics and idealism: the most obvious current example is the faltering state of sustainability-conscious investing, before it ever became properly established (the possibility of a looming end to the Trump era provides some fragile basis for optimism on that front at least). Perhaps the profession’s emphasis on itself as an expert in trust, however scrupulous, isn’t unrelated to the recurring concerns about recruitment and job satisfaction. Building trust may be important, but it’s sometimes in necessary opposition to the thrill of the new, to the embrace of risk, the conquering of unchartered territory, to the rush of pleasure that comes from running over the bedrock, rather than being embedded in it…

The opinions expressed are solely those of the author.