Despite AI investments, 88% rely on manual processes

Audit and advisory leaders are hard at work implementing AI solutions that they feel will be vital to future growth, but major implementation hurdles mean that most professionals remain largely dependent on manual processes. 

New data from audit solutions provider Fieldguide found a number of seemingly paradoxical points that show, on the one hand, a great eagerness to invest in AI solutions while, on the other, severely lagging behind on implementation. For example, while 94% report that engagements go faster when you have AI and 90% say AI reduces workloads, 88% still rely on manual data collection and 80% are still documenting by hand. Similarly, while 81% say AI has improved profitability, the same proportion report that rollouts were abandoned due to lack of stakeholder buy-in. 

Fieldguide believes this is reflective of the AI adoption deficit between organizations, which is how firms can run successful pilot programs but still struggle with the day-to-day. When firms do implement AI and do it well, they report higher earnings, reduced work and better retention. Once this is achieved, leaders say their roles shift more toward oversight versus execution (92%) and allows them to focus on strategic insights (81%). 

The challenge, of course, is getting there. Beyond internal pushback, other major barriers include poor data quality that leads projects to plateau due to evidence being scattered across systems and labeled inconsistently; and integration gaps, which makes it difficult to work AI systems into existing workflows versus existing beside them. 

The results call to mind the MIT study in May which said the vast majority of AI pilots produce no material impact on profit and loss, with only 5% generating significant returns on investment despite widespread adoption of the technology. The paper referred to the “pilot-to-production chasm.” This paper also pointed to lack of internal buy-in as a major barrier to AI adoption, though also pointed to the poor user experience and model output quality concerns. 

A recent report from industry analyst Gartner found that 25% of finance organizations are uncertain as to how best to transition from planning to executing AI solutions. Similar to Fieldguide’s findings, data literacy/technical skills and inadequate data quality/availability were cited as the largest obstacles to AI adoption across all organizations. Meanwhile, after overcoming the initial barriers to launching a pilot, it still takes time to realize significant gains, with 91% of respondents reporting low or moderate impact initially. However, the survey also found clear gains in organizations that did adopt AI successfully, as they were over two times more likely to experience moderate impact and nearly three times more likely to see high impact from the technology, and are half as likely to report low impact.

So how do you get from there to here? Fieldguide’s study found that those who successfully implemented AI solutions at their organizations had a few things in common. One was establishing an experiment-friendly culture, creating safe lanes for trials and retrospectives to build trust through incremental wins. Another was looking at data as a product, with successful adopters crediting standardized data and stewardship for their success. Finally, the vast majority codify lessons learned in playbooks and workflows, meaning AI adoption is a continuous process versus a one-off rollout. 

Given this, the report recommended that firms work to align their vision with their execution, understand implementation has a human as well as a technical element, unify systems and data to provide a strong foundation for AI, and redesign workflows for not just speed but sustainability as well. 

“The profession doesn’t lack belief in AI. It lacks follow-through,” said Jin Chang, co-founder and CEO of Fieldguide. “Audit and advisory leaders know AI improves margins and reduces hours. The challenge is turning pilot wins into firm-wide transformation. Those that close the adoption gap will define the future of the profession.”