FRIEDRICH-EBERT-STIFTUNG – Artificial Intelligence and Auto­mation in Retail 1 INTRODUCTION Digital technologies, automation, robotics and artificial intel­ligence(AI) are rapidly reshaping the global retail sector. The rise of e-commerce and omnichannel retail, the assembly of detailed consumer profiles through extensive data-gathering and analysis, and the intensification of work through surveil­lance and automated management systems, and many other applications are bringing significant disruption to the sector. The Covid-19 pandemic – while not a cause of changes in itself – has catalysed and accelerated many of these pre-­existing trends. This short report outlines recent technological developments in the retail that are linked to AI and automation and dis­cusses the implications for consumers, retail organisations and – especially – workers and unions. A core objective in identifying and examining digitalisation processes is to consider what role trade unions should play in the ongoing technological upheavals impacting the retail sector. The report summarises insights from five three-hour focus groups the researchers conducted with a cross-section of European trade unionists working in retail. The participants worked for unions affiliated with the trade union federation UniEuropa, which organised and facilitated the events. The focus groups included members from a diverse range of countries including Spain, France, Sweden, Denmark, Turkey, Italy amongst others. Focus group discussions examined AI, automation and robotics, the customer experience, sales work, and trade union responses in the context of digital technology adoption. Transcripts were analysed, summa­rised, and triangulated against publicly available data, quality academic studies and reliable media reports. The report first provides some contextual information about the sector and the drivers of the recent technological de­velopments. Second, it outlines and discusses some of the key technologies involving AI and automation that were highlighted by focus group participants as having particular implications for the sector. Third, the report discusses the potential role trades unions can play in responding to the challenges posed by technologies that make use of AI and automation. Finally, the report concludes by summarising the main challenges and recommendations highlighted by the research. 1.1 DRIVERS OF TECHNOLOGICAL DEVELOPMENT IN THE RETAIL SECTOR By way of background, the remainder of this section(1) briefly outlines the economic significance of retail alongside produc­tivity and employment trends in the sector; and(2) outlines some key technological trends in the sector, including the rise of e-commerce, and reflects briefly on the Covid-19 pandemic as a catalyst for change. 1.1.1 The economic importance of the retail sector and productivity trends The retail sector is an important part of the economy in ad­vanced economies. In most European economies, retail is the largest private-sector employer with the wholesale and retail trade employing 27.5m workers across the EU in 2019 – around 14% of the labour force. 1 The sector also indirectly supports many more jobs in areas like advertising, payments and logistics. In addition, the industry is critical entry-point to the labour market for many with 15% of European retail workers under the age of 25, and disproportionately high numbers of women and ethnic minority workers working in the sector. Despite its huge labour market significance, productivity in the sector is almost everywhere substantially lower than in manufacturing and other services. Retail did see a substantial increase in labour productivity from 1995 to 2006 as the use of ICT, logistics and supply-chain management technolo­gies – pioneered by the likes of Walmart in the USA – became generalised across firms and countries. These ›low-hanging fruits‹ of the digital and logistics revolution were largely ex­hausted by the mid-2000s. After 2008 retail sales fell due to the drop-off in demand initiated by the global economic crisis and from 2013 the recovery saw employers make growing use of cheapened labour – rather than continuing to invest in labour-saving technologies. As such, productivity-enhancing technologies developed some time back(such as self-scan checkouts) were not well-utilised due to unfavourable eco­nomic conditions and a reluctance amongst businesses to make large capital investments. At the same time, the low cost of labour and ongoing wage stagnation has enabled widespread automation-induced redeployment into low­value added jobs(e.g. from checkout operator to meet and greeting) – dragging down overall productivity growth in the sector. 2 2