Pressure on supply chains has never been greater as a result of prevailing consumer trends, shorter product lifecycles and the effects of COVID-19. These demands are accelerating adoption of supply chain 4.0 technologies and digital transformation. Robotics and artificial intelligence (AI), once the preserve of science fiction, are now a reality for retail and consumer goods supply chains. Technology is being used to automate and make decisions at every stage. Underpinning all of this, data is your crown jewel.
Retailers and consumer goods businesses can use AI and data analytics to drive predictive forecasting, where demand signals are used to determine levels for production and inventory, taking into account vast amounts of historic operational data, predictions regarding future trends based on ingesting real-time market reactions and environmental factors, such as weather patterns. Where any of these signals change, an automated supply chain will be able to adapt more quickly, mitigating risk. By predicting and reacting to changing demand levels, the supply chain is able to operate more efficiently, lowering costs and reducing waste and carbon footprint.
When considering implementing automated decision-making or AI, there are a number of legal risks to consider, even where, as is the case for most supply chain use cases, personal data is not being processed. Legal teams at retailers and consumer goods businesses need to:
- Ensure due diligence has been conducted on the underlying technology, so you understand how the tool will learn from your data and that the business understands how decisions will be reached.
- Appreciate what matters to your business from an intellectual property perspective – often the learnings cannot be separated from the underlying tool, so understanding how that improved tool will then be used by the vendor and its other customers is critical. Care should be taken if commercially sensitive data is being used.
- Check that you have the ability to use your data with the AI – this is a heightened risk where personal data is used, but consider any data belonging to the other actors in your supply chain.
- Examine auditability of the decision-making process – looking particularly at whether decisions can be traced where something goes wrong and making sure that accountability and liability is clearly allocated and understood.
- Mitigate the risk of any bias/discrimination caused by the data sets – for example, this could be relevant to supply chain, where AI is used to make decisions about product supply to the market.
- Understand how you get access to your data, the model parameters and the learned algorithms – this is key to avoiding supplier lock-in, should you wish to part ways with the technology vendor in the future.