The data isn't fit for purpose, but let's crash on with AI anyway! Workiva's global study exposes pragmatism and recklessness combined

Organizations which want AI to deliver competitive business advantages must concentrate on producing high-quality, trusted data, good AI governance policies, and role specific training.

That’s the not entirely surprising, but welcomingly pragmatic conclusion of Workiva’s recent Global Practitioner Survey, which polled more than 2,300 of said practitioners involved in financial and ESG reporting. It found that while 88% of respondents have seen ROI improvements in AI investments - 96% time saving, 94% productivity, and 92% cost efficiencies - two in three thought their companies currently lack the high-quality data, AI governance and training ready to use AI effectively. Despite this seemingly rather large shortcoming, the survey found that 74% of respondents are already using AI in their day-to-day work.

Today’s trade wars, supply chain instability, inflation, market volatility and regulatory changes are all adding to the strain on corporate financial reporting, but AI is, predictably enough given the hype cycle, seen by many respondents in the report as a way of coping with the increasing pressures. As per Kim Huffman, Workiva CIO:

Keeping up with AI technology is top of mind for executives. But what I found really interesting was that the survey found a high percentage of people are already using AI in their daily work. We are seeing increases in productivity, increases in return on investment, and an increased focus on higher value work for the individual.

Organizations that have taken a thoughtful, cautious approach to AI implementation have, unsurprisingly, a higher level of confidence in using AI, according to the report. Huffman suggests:

They've got a thoughtful approach, and organizations have focused on some really important areas to have a strong AI framework - data quality, data governance, responsible AI usage, AI literacy, and formal AI programs. They have technology professionals working closely with their business professionals to understand how to modernize their work processes using automation or AI. Alignment of how an organization is going to approach its AI program and its use of AI has become a Board level discussion.

The survey found that Sustainability professionals are more comfortable using AI than Finance professionals, with 64% saying that AI helps them to consume the large volumes of data they use, and 61% using AI for time consuming or manual tasks.  Huffman argues:

When you think about Sustainability, it's really about pulling your financial and your non-financial data together to drive insights for those sustainability professionals, so they're much more comfortable using large data sets for what they need to do, and the insights that they need in their roles.

She believes that reporting is no longer so much about compliance as offering a competitive differentiator, and AI is now being used to drive business insights and strategic decisions:

It’s become imperative that both the business and the technology teams work closely together to make sure they've got a tight environment for that kind of reporting. AI is so dependent on data quality, so pulling that data together into a centralized source is critical.

But, the data…

Many organizations that have begun AI implementations are finding data problems are now becoming evident. The report found that 64% of respondents thought their organization didn’t have AI-ready high-quality data in place, 65% said there was not enough effective AI governance, while 67% said there was no role specific training. As well as slowing AI adoption, these problems could expose companies to increased risk.

Keeping up with fast developing AI technologies is a priority, with 44% of executives in the survey saying adapting to new technologies is the top internal factor threatening performance this year. While on a day-to-day basis 42% said keeping up with technology or AI was the biggest challenge. According to Huffman:

For a financial reporting person, embedding capabilities that leverage AI in the applications that they already use increases the trust level. Additionally, the data that is already in that application, the data model, is already integrated, so you're not sending the data elsewhere, which poses data security, data privacy risks. You're confident that the data that's being used is staying within the application. I think it will give a higher degree of confidence that it's being used responsibly, and then the users are comfortable.

The report also found that organizations where AI is already used on a day-to-day basis are re-investing the time saved in business development. Nearly 40% said they were reinvesting their time in advancing sustainability initiatives, 36% to accelerate product or service innovation and 37% in improving customer experiences.

Workiva’s Huffman concludes that customers who take a cautious, pragmatic approach will see the benefits of AI adoption more quickly, without adding risk:

Where we enabled AI capabilities on our platform, we didn't default those capabilities. We worked closely with all of our customers to understand whether it is something they would like to take advantage of.  This enabled them to make the decision themselves. Security and data privacy, given the nature of our business, is paramount to us. So we took a very thoughtful and pragmatic approach to how we embed AI technology. We introduced capabilities and features slowly to our customers so that they can get comfortable.

My take

Using dirty data has always led to disaster. ‘Garbage in, garbage out’ is a phrase from the birth of the IT industry and it’s never been more true than today. AI will really throw a spotlight on companies which haven’t sorted their data out. Encouragingly, corporate financial reporting practitioners seem well aware of the risks, and the next 12 months should see a focus on improving data quality, and governance and security.  If it doesn’t…