Nicolaj Gudbergsen, Head of GenAI at Danske Bank, Karen Bodner, Head Global IR Advisory at BNY Mellon and Christopher Blake, Executive Director, Product Management, S&P Global discussed how far AI is implemented in business processes and how can it be used to enhance the efficiency and effectiveness of financial communication and other aspects of IR work. How to address the challenges, learnings and best practice of implementing AI tools to support IR.

Christopher Blake: The key takeaways thinking about how the buy side is using tools, is the data that is available on your company out in the public sphere. It’s more important than ever that it’s simple, concise, straightforward and not open to too much interpretation. The more that the data about your company and the things that you’re saying are open to interpretation, the more chances there are that AI tools could misinterpret them. And because people are using the AI tools instead of humans in some cases, you might open yourself up to situations where something is out there about your company that’s really not accurate, and you haven’t quite had the chance to address it because it didn’t come as a result of the human interaction.

Karen: “There are a lot of buy side firms, who are putting AI in their back office to automate their processes and that’s really the biggest thing, when we talk to investors. They’ say “that’s what we’re doing”, and it’s not front office; They’re not making investment decisions with it, but they’re using it to automate their processes. There is a subset of information and generative AI and tools that are out there that we think people need to be aware of, and the question is how you react against it.”

Nickolaj Gudbergsen on the use of imperfect AI models due to “hallucinations and biases”: “I have to stress and recommend that should not be a reason for not getting started, because everyone wants to solve these problems. There’s no one who wants hallucinations or biases, so don’t be too occupied on what we can do for you this month or next quarter, we won’t be certain about what can […] If you get behind the curve, or you maybe didn’t do anything and wait and see, then the catch-up race might be really, really hard because the other companies might already have been really embedded in their technology departments”.

Slido question 2: How are you using AI tools in your work today?

To a greater extent – I use AI tools regularly in my IR work compare to a yea a go: 66 percent

The same: 24 percent

What is AI?: 8 percent

To a lesser extent – I use AI tools less now than I did a year ago: 2 percent

Link to on-demand streaming of the session