For many people AI first appeared on their landscape with the release of ChatGPT last year. However, it has in fact been around much longer than that and has been in use by many Irish companies for years. In particular, machine learning (ML), a subset of AI, has been used by Irish businesses for decades to, inter alia, enhance customer care, sell more, and improve customer retention.
What is AI?
Let’s start by covering what AI is. There are many definitions, but they all evolve around the concept of computers mimicking human behaviour. More precisely I would argue that AI is software that identifies patterns in data and can regurgitate output based on those patterns. So, for example, an AI model could look at historical data on people and their credit card usage and learn from that what were the patterns of usage of those people who always paid their debt and the patterns of those who defaulted. Once the AI model has that knowledge it can look at a new set of data on credit card users and identify those who are likely to default in the future based on their usage patterns. This is exactly what many financial institutions do to minimise the risk of bad debt. For AI like ChatGPT, instead of predicting bad debt the model predicts which word is likely to follow another word or sentence. So, in this simplified example an AI model can predict the correct word because of the context.
Tom picked up the phone and said [ ]
Tom picked up the phone and said ‘goodbye’
Tom picked up the phone and said ‘how are you’.
Tom picked up the phone and said ‘hello’.
Even though AI models like ChatGPT seem intelligent they actually have no understanding of what they are saying, at least not in the way humans understand. However, they are nonetheless incredibly powerful and useful.
AI and Your Business
As we approach the end of 2023, AI is rapidly transforming the business landscape globally, and Ireland is no exception. The potential of AI to enhance operational efficiency, develop insights for product/service development, and improve customer engagement is increasingly being recognised by Irish businesses. However, the adoption of AI also brings with it a host of challenges and concerns, particularly around data privacy, security, and ethical use.
A recent survey by the Institute of Directors in Ireland revealed that only approximately 40% of organisations are using AI and, of those, a majority reported using AI to enhance operational efficiency. Over 60% of Irish directors/senior executives still do not use AI in any way in their organisations and a third of them have no plans to so - this is akin to someone in the 1930s saying they had no plans to use electricity in their business. To view this positively, this suggests a significant opportunity for growth and development in the AI sector in Ireland. Personally, I believe that AI will rapidly become non-optional for most businesses - it will be a case of adapt or die.
EU's Draft Artificial Intelligence Act
The IoD Ireland survey also highlighted a lack of awareness among directors about the scope of the EU's draft Artificial Intelligence Act. Over 75% of survey participants were not aware of the scope of the Act, and 45% of these had no opinion as to whether it is pitched at the right level. Our view in Idiro Analytics is that the legislation is long overdue and that it is a case of better-late-than-never. However, the technology will continue to evolve and so legislation and regulators will need to work harder to stay abreast of changes. What has been overlooked in the discussions regarding the AI Act is that it will cover computer aided decision making not just AI. So, if a company is using computer logic to differentiate between, say, customers then that logic (computer code) will be subject to the EU AI Act. When one takes that fact on board the reach of the AI Act spreads out like St Brigid's cloak covering an area much greater than originally anticipated.
The EU AI Act was originally proposed in 2021, the same year Idiro established its AI Ethics Centre, and is the first comprehensive set of regulations for the AI industry - though other jurisdictions are also working on similar legislation. The Act aims to ensure that AI systems are safe and respect existing laws, values, and rights. It also seeks to facilitate the development of a single market for lawful, safe, and trustworthy AI applications. The EU AI Act will, like GDPR, reach beyond the shores of the EU and any business that uses AI on the data of EU citizens will have to comply with the EU legislation.
From a business perspective, it means that many, though not all, companies that use AI will have to ensure that their AI models are fair and do not discriminate against anyone based on their gender, race, religion, sexual preference etc. So, for example, if a credit card company is using AI to determine who qualifies for a credit card, they will need to measure how many women qualify for their card and how many men. If it is found that more men qualify than women, then the credit card company will have to review their model and remove whatever is causing the bias. There are tools and techniques to do this which will have to become part of a company’s standard operations when developing and deploying AI models. The use of AI in the following sectors are going to be regulated by the EU AI Act but expect more to be added over time:
- Financial services – particularly in areas like credit scoring risk assessment
- Law enforcement
- Critical infrastructure
- Essential public and private services
- Access to education
- Administration of justice
- Biometric identification of people
- Migration, asylum, and border control management
The Act is currently now in talks with the EU countries in the Council on the final form of the law. The European Parliament has noted that the aim is to reach an agreement on the Act by the end of this year.
In light of these developments, it is crucial for company directors in Ireland to familiarise themselves with the EU AI Act and its implications for their businesses. They should also consider how AI can be integrated into their business strategies in a way that is ethical, lawful, and beneficial for their organisations. There will be winners and losers in the race to harness the benefits of AI and progressive directors will advocate for its adoption within their companies while being aware of the ethical considerations that need to be kept front-of-mind as this powerful technology disrupts, transforms, and enhances our economy and society.