In the rapidly evolving landscape of artificial intelligence, data governance has emerged as a cornerstone. As AI technologies continue to permeate various industries, the responsible and ethical handling of data becomes paramount.
However, as AI applications proliferate, along with the volume and sensitivity of data, a robust regulatory framework has emerged to address risks and ensure accountability. This is where businesses and other organisations today find themselves: they want to experience the benefits of AI, but they are naturally concerned about the associated risks.
Michael Cronin, co-founder and managing director of digital transformation specialist OpenSky Data Systems, emphasised that AI brought real value to businesses, but it had to be implemented with care.
“AI has always brought value, but today’s advances in generative AI, especially with large language models, allow businesses to achieve even greater AI potential when paired with robust data governance,” Cronin said.
But there is more to AI than getting a chatbot to write bullet points and the other tasks for which the likes of ChatGPT have been deployed, in many cases without authorisation.
“We apply large language models with precise scope, tailored to run on an organisation’s unique data. This allows businesses to address complex challenges specific to their operations, whether in finance, human resources, or operations,” he said.
Adapting AI to meet a company’s specific needs not only creates competitive advantage, it also means accepting the regulatory burden associated with processing data.
“Without data governance you can use AI, but it’s not going to be effective and it’s upping the risk in a major way. Governance is key for understanding the data you have,” Cronin said.
Indeed, recent years have seen data governance rise up the legislative agenda in Ireland and the EU as a whole. Cronin said companies needed to understand that this wasn’t going to change.
“GDPR hasn’t gone away, NIS2 is coming in, the EU AI Act is coming in and there’s even a new EU Data Act. That’s a lot of regulation to comply with.
“Without data governance you are really increasing the risk and, in addition, you are reducing the benefits of AI.”
OpenSky Data Systems did not just talk the talk, Cronin said. It has integrated AI and data into its own operations and uses its own platforms.
“Unlike many companies, we actively use the same AI and data solutions we develop for our clients, ensuring our expertise is built on real-world application.”
Without data governance you can do it, but it's not going to be effective and it's upping the risk in a major way. Governance is key for understanding the data you have
It is commonly accepted that larger enterprises have the advantage when it comes to AI, as they have more resources to invest. However, even a large enterprise could struggle without proper data governance, Cronin said, as they had a lot of data and a lot of systems, so, data discovery was an issue for them.
“In truth, these problems exist for SMEs, for large enterprises and for government. Of course, risk is a huge thing for the public sector,” he said.
Moreover, while AI adoption is on the rise, there has been a tendency to overstate how deeply it has been integrated into operations.
“What you have is reports saying things like ‘70 per cent of businesses are using AI’, and they are, but in many cases they are using publicly-available LLMs. The next step is to use the company’s own data in a proper and properly-governed way.” .
Ultimately, Cronin said, proper AI depends on data governance.
“Data governance is the cornerstone of AI success: it enables scalability, ensures responsible use and eliminates bias, positioning companies to fully leverage AI’s transformative power.”