Working from Home (Telework) and AI Decisions

The utilization of work-from-home (telework) has been recommended in major countries as a measure to control infections caused by the global COVID-19 pandemic. With employees also experiencing the benefits of telework, an increasing number of companies have made use of teleworking systems and actually telework has also been recognized as having a considerable policy effect to dampen the pandemic. Contrary to initial concerns, employees never "slacked off" with reports suggesting that they have been properly allocating their time to work. Even after the pandemic goes down, it is likely that telework will survive and continue to be utilized as an important work style.

Having said that, the management side cannot hide anxiety at not being able to directly observe employees working "on site." For this reason, many companies have introduced software to monitor the activities of employees during telework. Under current circumstances, it is common for employees to be made to submit daily reports on the details of their work on an hourly basis using something like Excel, in order to assess work efficiency, determine the appropriateness of their managers instructions and check for overwork or other issues from the view point of human resource management.

However, some companies are already working to develop highly sophisticated software using AI to identify employee aptitude based on the big data collected and attempt to make judgments on employee advancement and other decisions. It was around that time that the European Union (EU) released draft rules on AI.

According to the rules, AI systems that monitor the activities of people and use the results of that monitoring to make determinations on things like hiring, promotion and transfers are regarded as high-risk systems. In addition to a requirement that the people involved need to be familiar with how such systems operate in advance, the rules state that there is an obligation to comply with designated laws and regulations concerning those actions. However, that does not mean that no decisions made by AI systems can be trusted. As the decisions made by managers are possibly biased, it is expected that AI systems will be able to suggest better work styles.

But employees have a right to know the details about how these systems will be used and how they will be linked to their own performance evaluations. For example, just because an AI system says "no" does not mean that an employee's treatment has been adequately explained from a personnel perspective. If there was a manager who improperly forced through the results of decisions made by such a system, a company could become embroiled in a variety of labor disputes and lawsuits on the preservation of rights.

For the Ricoh Group, which advocates being a digital services company, it will be necessary to first adequately monitor the activities of our own employees engaged in telework and make persistent efforts to transform that experience into "Fulfillment through Work". Moreover, if we can utilize these experiences for our digital services business, we will also be able to contribute to the improved performance of many companies promoting telework, both at home and abroad.