The Implications of Generative AI in Retail M&A

With its ability to amalgamate and analyze virtually limitless sources, generative AI has the potential to transform the retail industry, how consumers shop, how retailers do business, and how retail businesses are themselves bought and sold. But what is generative AI? How can retailers best leverage its potential and what does the adoption of this technology mean for retail M&A activity?

Generative AI, also referred to as machine learning or large language models, are algorithms that can be prompted to create content, including text, code, images, audio and video based on what they’ve gleaned from materials they’ve ingested. Applications for generative AI in the retail sector include streamlining and automating customer support (including customer service chatbots, AI-powered shopping assistants and personalized product recommendations), improving supply chain processes such as forecasting demand and managing inventory, and monitoring and optimizing pricing to lend competitive advantage. Retailers may also employ generative AI in connection with their marketing activities, for instance to improve product pages for e-commerce sites and to maximize the value and placement of ad copy and imagery. For brick-and-mortar retailers, generative AI has the potential to unlock value in terms of selecting new retail locations that are most likely to be profitable and informing retailers as to how to optimize existing store layouts and operations.

Product designs are also increasingly impacted by generative AI technologies. We’ve seen retailers digest large swaths of customer data to identify trends and features most likely to increase sales, and to use generative AI to design and manufacture products to fill those identified gaps and meet predicted trends.

Finally, generative AI has the power to detect counterfeit goods and fraudulent activity, allowing retailers to mitigate risk in these areas and achieve savings.

While the potential applications of generative AI in the retail sector are plentiful, as are their prospective benefits, there are a number of legal considerations to be aware of when using these technologies. This is particularly true in the context of intellectual property ownership issues, accompanying risk and associated liability.

First, while generative AI has the potential to create useful and valuable output, the ability to hold intellectual property rights in that output is presently unclear. Current Copyright Office policy takes a bright line position that AI-generated work is not the product of human authorship, and so cannot be copyrighted. This approach subjects AI-generated designs, marketing materials, copy and imagery to use by others including competitors, potentially without legal recourse. And while the Copyright Office did issue a Notice of Inquiry to the public on this topic in late 2023, it is unclear if or when this policy will shift.

In addition to ownership-related considerations, the use of generative AI in the retail sector can also implicate potential liabilities. Because of the way generative AI is trained, arguments have been advanced, including in the context of pending litigation, that its output is derivative and thus infringing of the copyrighted materials on which it has been trained. This means, for instance, that where a retailer uses generative AI to design a product, and the generative AI model has been trained on a third party’s designs (assuming, of course, there are protectable features of those designs) the retailer could be subject to IP litigation depending on the output generated and utilized. There is also the potential for vicarious liability associated with AI-formulated user-generated designs for customizable goods, as well as product liability considerations in light of generative AI’s propensity towards bias and hallucination, i.e. error.

We are seeing potential acquirers of retail businesses consider a target company’s use of generative AI, and the impact of such consideration in the retail M&A market. The due diligence process from an intellectual property perspective centers generally around ownership and infringement – what a target company’s IP portfolio consists of, and the risk of third-party litigation arising out of the IP owned and used by the target company. Intellectual property due diligence request lists are increasingly being tailored and refined to assess a target company’s use of generative AI so that the potential risks attendant to such use can be quantified. The target company’s use of AI is also becoming more relevant in an assessment of the company’s overall IP risk profile. These diligence requests can be nuanced in order to squarely address how generative AI is used by the retailer, the nature of the AI being deployed, and its origins. generative AI-focused diligence.

To mitigate risks identified during diligence, buyers of retail companies are starting to require more specific representations and warranties around generative AI (and require indemnification obligations for breach of such representations). There have been a variety of representations and warranties along these lines.

We’ve also seen an uptick in requests for representations and warranties relating to the target company’s compliance with applicable laws relating to AI technology and assurances that the target company has implemented and complied with industry-standard safeguards regarding use of AI technology. In the infringement context, buyers are requesting AI-tailored non-infringement representations and warranties stating that the company’s use of AI has not limited its ability to use or commercialize its intellectual property or products, and that such use does not infringe third party rights.

Where due diligence uncovers material risk areas with respect to a target company’s deployment of AI technologies, this can impact valuation and potentially dissuade a buyer from concluding the acquisition. On the other hand, successful deployment of generative AI to optimize sales, streamline internal operations and attract new customers can positively impact a retailer’s bottom line. The key to leveraging generative AI’s benefits in the retail sector while mitigating associated risks may thus lie in a bespoke approach, threading together various policies, practices and controls into a tailored solution designed to fit the needs, circumstances and use cases of each retailer. And as generative AI becomes more closely ingrained in the functioning of a retailer’s operations, we expect to see the impact of generative AI more prominently as companies trade hands in the M&A market.