“It is undisputed that AI has the potential of a so-called basic innovation,” emphasizes Dr. Eric Scheidegger, Head of the Economic Policy Directorate and Deputy Director at Swiss State Secretariat for Economic Affairs (SECO) in Bern. Nevertheless, he believes that breakthroughs in these technologies with a broad impact are likely to lie in the distant future: “Social, economic and legal considerations have always set limits to what is technically feasible.”
Taking into account employees’ fears and concerns
Currently, AI projects are mainly used to automate business and IT processes. In most cases, the aim is to automate standardisable routine tasks that do not necessarily require human intervention.
Classical examples are invoice auditing, the auditing of compliance and fraud cases or claims management in the insurance industry. However, chatbots also handle standard customer enquiries on telephone hotlines or inbound support in IT service management.
Will the machine relieve or displace people in the near future? Companies will have to face the fears and concerns of their employees. They have to develop concepts on how they want to deal with their workers in the future – and how they react to their questions about employment and the workplace.
Whether, where and how activities are to be structured differently or even new employment opportunities created, all these places high demands on personnel planning and development. The successful introduction of AI solutions, therefore, requires successful change management within the company.
Further obstacles need to be overcome
But there are a number of other obstacles that play a role and need to be overcome.
Data protection: According to a study by the market research company Lünendonk & Hossenfelder, the most frequently mentioned obstacle is the required protection of sensitive customer data. There seems to be a particular concern in the retail and logistics sectors, where around a quarter of all respondents indicated this obstacle. Just as many also feared that the use of AI would lead to new security gaps in the IT systems.
Data quality: However, more than half of the study participants, around 60 per cent, also consider the quality of their data to be insufficient to introduce AI applications for the analysis of data stocks or to enable machine learning. “Data quality” is seen as an obstacle, especially by companies in the telecommunications and retail sectors.
Excessive technological complexity and a lack of specialist know-how: this, too, was frequently cited in the study as an obstacle to the rapid implementation of AI projects. Although many companies are now dealing with this topic, there is still often a lack of experience to successfully tackle larger projects. In addition, many companies lack a distinct culture of error in order to try out new things without being afraid – as the participants in the study realised.
Lack of management support: What the AI experts in the company advocate and forecast does not always meet with the full acceptance of the management. This is why concerns about necessary investments or lack of time and resources are often cited as stumbling blocks.
For the introduction of AI solutions in the company to work, leadership is key: it is about finding committed people within the organisation who have the necessary enthusiasm, a budget and are willing to actively address the issue. This is best achieved by solving problems that not only affect individual business units but benefit the entire company.