Artificial intelligence may not be as powerful and dangerous as many people think or fear but care still needs to be taken with its deployments. This is the view of Dr Stephanie Dossou, director of AI at Mazars. “A lot of myths have grown around AI, and these have come from ideas of AI born out of sci-fi and pop culture,” she says.
She explains that AI is made up of a broad set of different technologies, which can be defined and grouped as self-learning and adaptive systems. One example is a real-time website chatbot to respond to customer queries.
“AI can bring large amounts of data together quickly and help us to identify patterns we wouldn’t be able to see on our own,” she adds. “It can help test different hypotheses and develop new solutions to problems.”
One of the greatest fears in relation to the technology is that it will replace people’s jobs, but this may be unfounded. “A government report published earlier this year on future skills and AI in Ireland indicated that AI is not likely to bring about a net loss of jobs,” Dossou notes. “It is expected that many jobs will change, with specific tasks taken over by AI systems. Other tasks more suited to human intelligence will continue to be done by people. It also noted that AI has the potential to bring substantial productivity increases.”
Generous State supports for R&D available as Enterprise Ireland grants and tax credits
Passing on a family business can be hazardous for even the smallest of businesses
Business energy supports ‘very significant’ for firms struggling with costs, say Mazars
Transfer pricing can be an easy target for tax authorities
She emphasises the need to keep humans ‘in the loop’. “There should always be a human included in the decision-making process using AI, no matter how advanced the AI agent is. With the augmented intelligence approach, humans are assisted by AI models with insights and recommendations to help them make more informed decisions, but the ultimate decision is always made by the human.”
Her role is both internal and external facing. One example of her internal work is supporting the audit department. “A lot of auditing is very repetitive, we do similar things year after year, and you can automate a lot of it using technology. The implementation of AI can help to deliver more effective and higher-quality audits. It also offers the auditors a more engaging and rewarding work experience.”
She also assists Mazars clients with the implementation of AI. “AI can be applied to many other sectors and industries, such as retail. Customer service can be transformed using AI-powered chatbots to strengthen customer service and respond to queries. AI tools with empathy are being developed at the moment. They can identify a customer’s mood and respond in appropriate terms.”
Dossou works with clients to deliver AI solutions that support diversity and equality. “For example, Mazars has a tool that can be used to measure remuneration and identify the gender pay gap within an organisation. Our tool brings this data together to assess trends in gender pay at all levels within an organisation. This allows HR and payroll teams to focus on anomalies, investigate discrepancies and implement strategies to address gender pay gaps.”
That is an example of AI as a benevolent technology. However, it can have adverse outcomes. “Some decisions made by AI can seem strange, and they are very often based on black-box models,” Dossou explains. “We just don’t know how the decisions have been made. That’s where explainability and ethics come into it. For example, loan applications may be judged consistently based on the criteria set, but there may be exceptions that it will not take into account. That’s why you need the human element in the process. AI can have biases and is not objective. It will depend on the data sets used, and how it has been trained and programmed.”
Progress is being made in this regard. “I am a judge for the AI and ethics category at the upcoming AI awards. It has been really interesting to see how companies in Ireland operating in different industries are developing and implementing ethics in AI. ”
“Overall, there are many benefits to AI,” she concludes. “Organisations connecting AI with their company’s mission will likely reap the rewards. For example, companies can adopt AI to add value to the customer journey. Support functions within an organisation may implement AI to assist with repetitive tasks and focus on value-added activities, which can also help in a tight labour market. It is essential to have a good governance model and strategy for AI and allocate appropriate resources to ensure that companies can successfully leverage the benefits of AI and ensure transparency and explainability in the decisions it makes.”