A new book gives concrete AI insight support to boards
TEXT: INGRID KINDAHL | ILLUSTRATION: RONJA MELIN
Article Originally appeared in The Swedish Academy of Board Directors Members Magazine “Professional Board Work” Nr 3 2020 (Professionellt Styrelsearbete)
AI will make humanity more intelligent. But it will require a significant effort and competence increase. Fortunately, there is now a book that helps boards structure their AI work.
Many board members have a high distorted self-image when it comes to their own digital and AI skills. According to international research, 62 percent of board members surveyed state that they are “savvy,”, ie, smart and knowledgeable, about digitization and AI. External evaluations, however, show that the real figure is 24 percent.
At the same time, research shows that companies with digitally competent boards have 38 percent higher sales growth than others, 34 percent higher return on total capital (ROA), and 34 percent higher market growth. It also shows that three AI-competent members of a board are required to achieve an effect.
Released this summer
All this, and much more, can be read in the book “AI Leaderships for Boards – The Future of Corporate Governance“, which was released this summer. The book has come as a result of a Vinnova-funded research project, led by Professor Robin Teiglandfrom Chalmers together with the professional board member, researcher, and advisor Liselotte Engstam and researcher Fernanda Torre at Innovation House HHS. Also, several other international researchers have participated. It has also collaborated with, among others, the large companies’ digital partner company Combient, the mining company Epiroc and the consulting company FCG Group, where Liselotte Engstam is chairman.
Well known to the members
Liselotte Engstam is well known to the members of the Swedish Board Academy, especially as a teacher in the education “Master focus in board work: Higher Chairman Training ”. Liselotte is herself a Civil Engineer from Chalmers University of Technology and has worked for many years with digitization, both within boards and in other ways.
– There are two significant misconceptions among board members about AI. One of the misconceptions is that it is only about technical approaches and concepts, such as deep learning, which scares many. The second is that many believe that AI should support the boards making their decisions. But it is still a long way off, says Liselotte Engstam. Instead, it is more about company development, she explains; innovations, business ideas, competition, and productivity growth.
The research work itself, which builds on both interviews and quantitative surveys with Swedish and international corporate boards, gives a good picture of what boards can and cannot do and what they do and do not do. But it is the book that will probably be perceived as most useful for boards. It provides hands-on help in structuring the board’s AI work. In words, tables, and checklists, concrete advice is given on what board members need to do. After reading the book, you can just get started.
In short, the board needs to focus on two significant areas: guiding company management and monitoring. The guiding can, in turn, be divided into three areas:
- Start with the company data. According to Liselotte Engstam, most companies have less order in their data and no strategy that outlines which data is most important in the future.
– What you often have is the current data, for example, what and how much you sell. But soon, it may be essential to know how the products are used and how they can be combined with other organizations’ data, for example, about the weather. Then you need to find your new required data, collect it and possibly form an external partnership around it, says Liselotte Engstam.
- Where should the company apply its AI innovation? Should it be targeted internally to improve your own business, or should you incorporate AI into your customer offering? Here, the board must guide.
- How should the company develop the ecosystem that arises through the use of AI? New types of business logic will require more collaboration between companies.- Like this with the weather – From whom are we going to acquire that data? Do we dare to choose a small startup that risks collapsing? Or should we buy a company? If our partner becomes listed on the stock exchange, will we lose control? So far, boards have mostly focused their own company – now new strategies are needed, says Liselotte Engstam.
The monitoring part, the board’s additional primary task, also has three points:
- Ethics, including what applies to the so-called black box. AI is often based on machine learning based on massive amounts of data. A product or result is produced without people being able to explain what data was used and how.
– Let’s take an example: HR departments can use recruitment tools based on AI, which is based on old data. If all those recruited are white men in their 35s, then the tool learns that we want to hire white men in their 35s, says Liselotte Engstam. Management must be able to go back to the recruitment tool (company) and express that this is not good enough. The board must set guidelines and policies for, for example, ethical results and monitor it.
- Cybersecuri There are two dimensions here: 1. When so much is digitized in the black box, can something undesirable enter the tool? 2. AI needs to be applied back to find attacks and stop them.
- AI will come in through other companies as it forms an ecosystem for all data. Then your company must be able to take responsibility for others’ AI as well. People need to take the lead in digital ecosystems not to fall behind and no longer be able to control.- Until now, the board has only had to think about algorithms that the company itself has developed, but now it must have control when using other company’s algorithms. If your company uses discriminatory tools, you cannot blame those who have developed the tools, says Liselotte Engstam.
Half have no plan
Half of all company boards today have no plan to either guide or monitor the development of AI. Few have strategies for how to use AI to develop new products and services and improve existing ones. But the research project has also identified a positive factor: Chairman is generally more ambitious than the boards.
– It is excellent because the chairman focuses on the work of the board. They can put AI development on the agenda and help ensure competent people on the boards. One way is to create technically-oriented committees that can help. This is what we have done when it comes to audit committees, and there is nothing to prevent you from doing the same thing with technology, says Liselotte Engstam.
The debate about AI tends to focus on ethics. Still, Liselotte Engstam points out that we must also work with development and innovation, which could be facilitated in technology committees.
Of course, AI will also be used by the boards themselves for their internal work. It can be about recruitment and evaluation of the board work and simple tasks such as transcribing voice during the meetings. Eventually, AI will be used as a basis for decision-making, and robots will undoubtedly participate as “board members”. – But we are not there yet, says Liselotte Engstam.
The research project, which has been going on for two years, is based on interviews and several significant surveys. One repeat survey was done by the advisory company Digoshen, conducted among approximately 400 board members from the Swedish Board Academy, ten more European countries and the USA, Australia, China, and countries in Africa. The first survey was conducted in 2016 with a follow-up 2017–2019.
Here are some of the results:
- In 2016, only 30 percent stated that they had a vision for the company’s digitization. The figure increased to 73 percent in the second survey.
- The use of digital technology to better understand customers increased from 40 to 58 percent.
- The proportion of companies that started sacrificing revenue to invest in digitization rose from 25 to 47 percent.
- 14 percent had begun to explore the next wave of digital opportunities, such as Al, robotics, and 3D printing in 2016. However, the figure rose to only 19 percent.
- In 2016, more than 50 percent answered that the company’s CEO had taken essential steps strategic digital business initiatives. The figure rose to 73 percent.
- In 2016, 25 percent had clear roles and division of responsibilities for digital initiatives. This figure improved, but only to 36 percent.
- In 2016, 60 percent of companies were not aware of the digitally related risks. The proportion remained unchanged.
- 66 percent of the board members claimed in the second survey to listen to customers, employees, partners, competitors, and industry experts via social media. This was an increase from 50 percent in 2016.
This is how AI can change the work of the board.
AI means new demands on boards and their board members.
Here are six points that will be needed.
- Improved balance between support development and control. At present, most of the board’s work
- is used for control.
- More dynamism. Traditional board work must be supplemented with faster adjustments of strategies.
- Better external monitoring. By keeping track of relevant indicators from business and industry, decisions can become more fact-based.
- Extended focus. From a narrow focus on shareholders to a broader view that encompasses allstakeholders.
- Greater interest in ethics and trust. Companies need to raise their ethical standards. It is no longer enough to follow the law.
- Both short and long perspective. Boards will need to see the company’s future in both the long and short-term views and combine scenario thinking with strategy development and implementation.