s software agents to take proactive actions, similar to humans. Coresight Research estimates that the global GenAI applications and hardware market will total $125.0 billion, $39.6 billion for applications and $85.4 billion for hardware in 2025. Additionally, the advisory firm calculated that year-over-year (YoY) growth for the total market will reach 57 per cent in 2025, marking a major expansion of GenAI use, especially in regard to GenAI applications, which are expected to see 106 per cent YoY growth in the same year. While it’s not too late for retailers and brands to explore AI use across various functions, John Harmon Coresight’s managing director of technology research cautioned, it is clear that the leaders in this space like Target and Walmart, which have already implemented GenAI company knowledge bases, are strongly ahead of the curve. 2025 is expected to be a pivotal year for the development of AI, with many GenAI projects seeing deployment, so retailers need to shape up or ship out when it comes to the AI curve. Top 10 AI trends to tap into in 2025 In a report titled “Retail 2025: 10 AI Trends- An Inflection Point in the GenAI Revolution”, Coresight Research revealed 10 recommendations for retail organisations to harness the power of AI in the year ahead: 1. Untether retail media Retail media is clearly a high-margin revenue opportunity for retailers, one that allows brands to leverage retailers’ first-party data and reach consumers closer to the point of purchase. Coresight estimates that the US retail media market will reach $67.8 billion in 2025, ultimately growing to $106.4 billion in 2028, representing a 2025–2028 compound annual growth rate (CAGR) of 16 per cent. However, Coresight analyst Charlie Poon warned, while retail media is a fast-growing sector, there are several challenges to its growth. Challenges include fragmentation in the ad buying space, an inability to demonstrate return on ad spend (RoAS), limited ad inventory and the inclusion of various manual tasks, such as designing and launching ad campaigns. It will be vital for retailers to tap into AI machine learning (AI/ML) to make retail media networks more efficient and powerful in the year ahead. Three key use cases of AI/ML in retail media centre on: Audience segmentation: AI can draw insights from a vast number of data sources, while advances inGenAI enables data cleansing and drawing insights from structured and unstructured data without the need for complex algorithms. AI/ML can also help identify patterns in customers’ shopping behaviours, including key triggers to purchase, allowing for improved RoAS through increased relevance. Data utilisation: AI can be used to merge real-time insights on shopper behaviour with predictive analytics, helping brand advertisers optimise their ad campaigns and build targeted, personalised campaigns using contextual data on shoppers and products. Ad buying: AI can substantially elevate both self-serve and programmatic ad-buying infrastructure through its ability to ingest millions of data points within seconds, helping media buyers select the optimal ad format, including the time and location the ad will air. 2. Catapult product development to a new level GenAI offers several ways for brands to accelerate and expand their creativity with new product development processes, especially given its ability to analyze and summarise text. For instance, GenAI models are able to mine social media posts for significant oremerging customer trends, which can then be input into image-generation applications fornew product ideas. Correspondingly, the models can analyze product reviews for new product ideas, as well as digest customer feedback to identify quality or sizing issues and enhance product descriptions. 3. Accelerate product description and imagery creation Detailed, accurate product descriptions and imagery are an unavoidable task for any digitally-based retailer. However, it is one that comes with several challenges such as sparse or inadequate information from supplies or just the sheer volume of information needed when multiple products are added to a store at once. Gen AI can accelerate the time-consuming task of creating product descriptionsand imagery, while large language models (LLMs) can understand the context and semantics of the content, allowing retailers to generate detailed and engaging product descriptions that align with customer preferences. Furthermore, by incorporating a retailer’s proprietary data, such as its brand voice, Poon pointed out that the retailer can use LLMs to automatically generate product descriptions that not only accurately represent the product, but also the company culture. 4. Unlock data access and analysis for all Until relatively recently, enterprise data was limited to the rare few, such as data scientists—who understood the tools necessary to extract insights from this wealth of data. However, through the use of Gen AI, data can be interpreted in an easier-to-understand language for the everyday employee. This in turn, allows everyone across a company to ask data-related questions in real-time and in an interactive manner, often eliminating the need to submit requests to a data-science team and greatly speeding up the process of finding insights. For example, retailers and brands such as Target, Tropicana and Walmart have already implemented these types of knowledge bases, which allows their associates to be more productive and spend more time with customers. 5. Create a new workforce with agentic AI Just a little over two years after the debut of ChatGPT in November 2022, the retail industry is now experiencing a new wave in AI development, leading to the creation of AI agents that are dubbed agentic AI. These AI agents are instructed in natural human language, similar to the written instructions one would give a co-worker, taking AI democratisation to a new level. They then use GenAI to understand what is needed and activate software APIs (application programming interfaces) to carry out that action. For example, AI agents are able to launch a marketing activation or send a customer a follow-up email. The retail industry is still in the early stages of discovering all that can be done with AI agents. At the AI, Agents and Applications conference in November 2024, renowned computer science professor Andrew Ng defined four stages of AI agent implementation: reflection; use of tools; planning; and multiagent collaboration—this revolutionary fourth stage features agents interacting with other agents, which would revolutionise the potential productivity gains and accelerations of AI. 6. Level up to demand forecasting 3.0 “Gone are the days when retailers could simply use past sales data alone to predict future demand,” Harmon emphasised.It has become increasingly difficult to accurately predict demand thanks to factors like evolving customer expectations, influence from social media and economic and political fluctuations, as well as supply chain disruptions and returns that can severely disrupt demand forecasting and inventory management. This is where AI/ML comes in for the win, as this technology excels in ingesting multiple sources of data and generating an accurate forecast. Additionally, AI-powered automation features can enable a demand forecasting platform to act, and emerging AI agents will take this ability to act one step further. According to data collected by Coresight Research, bringing in external data, such as weather or calendar events, can boost the accuracy of demand forecasts further, from an estimated 50 per cent to 90 per cent or higher. 8. Bridge the gaps in fragmented supply chains One of the many issues highlighted by the pandemic is the severe gaps of information that are caused by the many fragmentations in the supply chain system. As supply chains are a patchwork of factories, sensors, ships, planes and trucks with no common language, interface or data standard, it means that the data readiness of each of thesepoints vary greatly and can lead to major delays in the production cycle. One way to work through these gaps is through the use of AI. For instance, traditional AI/ML can estimate transit and delivery times, as well as monitor products within the supply chain, generating alerts when target thresholds are missed. AI can also present data and estimates in easy-to-understand dashboards, as well as assess signals and messages from the supply chain, handling non-urgent issues and escalating issues that require immediate or human action. 8. Fulfill the promise of personalisation at scale Realistically speaking, it is nearly impossible for a human team to enable hyper-personalisation at scale. By incorporating Gen AI, which will ingest and aggregate extensive amounts of consumer data in real-time to generate customised communications via a variety of channels, including chatbots, websites, e-mail, social media and text messages, hyper-personalisation at scale can be achieved. This greater relevance of communication will in turn drive customer sentiment, and therefore, conversion rates and revenues for retailers. 9. Deliver real results with contextualised search The Coresight Research team predicts that consumers will growingly lose patience with traditional keyword-based search and increasingly turn to Gen AI-powered search, making it crucial for retailers to invest in the tech if they haven’t already. Additionally, thanks to the development of new, AI-driven search features, such as visual and voice search, consumers will increasingly tap into Gen AI-powered searches thanks to its enhanced convenience. 10. Empower workforces with AI Retailers currently face a host of challenges in enabling associates to offer intimate service to consumers. However, the potential of AI/ML and Gen AI promises to enhance associates’ capabilities,helping to offset these challenges. As Coresight’s managing director of technology research explained, “AI can be used to offer various features that empower associates and enhance the front-of-house experience.” This includes enhancements like personalised recommendations or promotions that take into consideration a customer’s preferences and past purchases to virtual try-on capabilities. The opportunities for AI-powered growth within the retail industry are boundless. However, the Coresight Research team warned retailers and brands to consider their uniquebusiness needs before jumping into large AI projects. Starting out with small bits of experimentation can often be beneficial and will eventually allow companies to build up to larger AI projects, the advisory firm cautioned.