Despite a growing recognition of the potential of artificial intelligence (AI) to revolutionise industries, a recent survey by technology consultants BearingPoint revealed that only a fraction of Irish businesses have taken concrete steps to implement AI technologies.
According to Stephen Redmond, director and head of data analytics and AI at BearingPoint, this strongly means that while awareness of AI’s transformative power is widespread, the practical application of these technologies remains limited in the Irish business landscape.
“Our survey said something like 19 per cent were [already] adopting AI technology and 47 per cent are still thinking about the best approach to governance and decision-making for AI,” he said.
In addition, some 32 per cent of organisations surveyed reported having no AI implementations or not having yet started considering AI technologies.
A key roadblock is data, which companies know they need to get right: some 82 per cent of C-suite executives responded that they saw establishing effective governance as a crucial step in becoming an AI-augmented organisation, while 35 per cent of organisations report a lack of readiness in terms of data availability and quality for generative AI projects.
Putting people in the decision-making role with AI is important
They are not wrong: data does need to be available, it needs to be prepared for AI, and it needs to be processed in a compliant manner. However, the delay in adopting the technology raises the spectre of the potential for unauthorised use of AI by staff – so-called ‘shadow AI’.
“Among the laggards you can bet there is shadow AI,” Redmond said.
As BearingPoint is a global organisation, its survey went out to 700 organisations right across the globe, mostly in Europe but also in Asia and the US. Of those, 170 were in Ireland and Britain. The objective, Redmond said, was to see what the current state of affairs was, and to enable BearingPoint to help move its clients forward to allow them to use AI to meet their goals.
“We are talking about what people need to do, what are the leaders doing to get ahead, what are their people saying,” he said.
While governance may act to slow adoption, it is crucial that businesses take it seriously, as failure to comply results in the potential for reputational damage and eye-watering fines.
“Some will have very good governance, which is very important, especially in light of the EU AI Act,” he said.
The AI Act, in addition to other relevant legislation such as the Data Act and of course GDPR, straightforwardly sets out to ensure that data is not misused, for example ensuring privacy as well as not replicating biases that could have an impact on decisions made in areas such as credit provision.
“It’s about protecting consumers and it’s putting the onus on the people who make AI, but also on people who deploy them, to ensure they are appropriately assessing the risk,” he said.
Crucially, this means that while getting deep value from AI means working with relevant data, it needs to be suitably controlled.
“The data you’re now using is what you need to be able to deploy AI to process, and do it in such a way that the AI model doesn’t know about your internal documents,” Redmond said.
“There are things you can do with AI, with specialists who will come in and clean up your data. However, if you have documents sitting on a SharePoint and you don’t know what’s in them you can get an AI to classify them. That is very likely to be useful to a wide array of organisations,” he said.
Culture is going to be a huge differentiator in AI adoption, Redmond said, and organisations that allay fears around AI will see more progress.
“Putting people in the decision-making role with AI is important, and it is also a way to address the resistance,” he said.
“Fear of losing a job is natural, but there are a lot of opportunities. I heard a great quote that I think sums it up well: AI won’t replace your job, but a person using AI could replace your job.”
Ultimately, Redmond said, these social aspects of AI are the parts that organisations most need to pay close attention to.
“Organisations have gathered a lot of data over the years, so there is a lot of work to do. The technology is the easy part: it’s the governance, the people management and the data management that are the big three,” he said.