AI shopping adoption accelerates ahead of holiday season as trust emerges

Nearly half of UK adults now use artificial intelligence tools regularly, and almost a quarter plan to deploy them for Black Friday and Christmas shopping according to research released November 25, 2025 by PSE Consulting. The nationally representative study of 2,000 UK online shoppers reveals 85 percent of those planning to use AI for Christmas shopping would trust it to place orders and execute payments on their behalf.

The findings arrive as McKinsey research projects that agentic commerce—shopping powered by AI agents acting autonomously on behalf of consumers—could orchestrate between $900 billion and $1 trillion in revenue in the US B2C retail market alone by 2030. Global projections reach as high as $3 trillion to $5 trillion, according to the consultancy's November 2025 report.

AI adoption demonstrates particularly strong penetration among younger demographics. Among 18-34-year-olds, 42 percent intend to use AI for holiday shopping, highlighting rapid uptake among younger, urban and more affluent consumers, according to PSE Consulting. This contrasts sharply with older demographics, where more than half of those aged 55 and above never use AI tools and 80 percent will not rely on them for festive shopping.

Chris Jones, Managing Director at PSE Consulting, characterized the shift as fundamental. "What's remarkable is how quickly AI has moved from novelty to an integral part of shopping," Jones stated. "AI is now becoming an active participant in the payment process. Systems designed for human-paced transactions are now under pressure to support high-frequency, autonomous agent-initiated flows."

The payments infrastructure faces significant structural challenges as commerce transitions from direct user interactions to agent-initiated transactions, according to McKinsey analysis. Components such as gateways and fraud engines have been built around a human-in-the-loop model, in which identity, intent and authorization are explicit and observable. Agentic commerce disrupts this paradigm: the "customer" is now an AI agent acting on behalf of a person, necessitating new approaches for delegated authorization, programmable spend policies and consent attestation.

OpenAI and Stripe launched the Agentic Commerce Protocol on September 29, 2025, enabling transactions from Etsy and Shopify merchants directly through ChatGPT conversations. The service is expected to reach Europe within six to nine months due to issues with Strong Customer Authentication, according to PSE Consulting's research findings.

Google launched Agent Payments Protocol in September 2025, a payment-agnostic protocol designed to empower autonomous and semi-autonomous AI agents to make verifiable purchases on behalf of users. The protocol utilizes cryptographically signed mandates that link intent, cart and payment across users, merchants and payment networks, creating an audit trail not subject to repudiation.

Payment industry leaders are pursuing parallel innovations. Mastercard is developing its Agent Pay solution, while Visa is positioning its global network as the backbone of agentic commerce. In partnership with AI platforms including Anthropic, IBM, Microsoft, Mistral AI, OpenAI, Perplexity, Samsung and Stripe, Visa is piloting transactions where AI agents can spend on behalf of users within preset budgets and consent parameters. Visa has also launched AI-ready cards, which replace static card details with tokenized digital credentials.

Cloudflare announced on October 24, 2025, a collaboration with Visa and Mastercard to develop security protocols for automated commerce as artificial intelligence agents begin purchasing products on behalf of consumers. Visa developed the Trusted Agent Protocol while Mastercard created Agent Pay, both leveraging Web Bot Auth as the foundation for agent authentication.

Despite the rapid adoption, UK consumers remain cautious. Privacy, fraud and accuracy concerns are significant barriers, with 49 percent worried about data handling, 46 percent about fraud and 41 percent fearing AI might select the wrong item, according to PSE Consulting. Only 9 percent report no concerns, indicating that trust and transparency remain critical for widespread adoption.

The PSE Consulting research identifies an emerging "AI shopping confidence divide." Early UK adopters—typically aged 18-34, affluent and urban—use AI daily or several times a week. They are almost twice as likely as the general population to plan AI-assisted holiday shopping, though they remain cautious about practical issues such as returns, incorrect items and opaque checkout flows. At the other end, UK traditionalists, typically aged 55 and above with lower digital exposure, remain hesitant.

Jones added that the emerging divide is reshaping shopping behavior. "Early adopters are already turning to AI daily to hunt for deals and select gifts, while more traditional shoppers are taking a wait-and-see approach," he stated. "Businesses that want to stay relevant will need to rethink how they engage customers, whether through integrated AI tools, personalised assistants, or new ways of presenting products that appeal to both early adopters and more hesitant consumers."

McKinsey research indicates that agentic commerce is taking shape through three key interaction models: agent to site, where agents interact directly with merchant platforms; agent to agent, where agents transact autonomously with other agents; and brokered agent to site, where intermediary systems facilitate multi-agent and multi-platform interactions.

The rapid pace of innovation in agentic AI is being driven by six key tools and developments, according to McKinsey. Model Context Protocol allows AI agents and systems to share context, intent and data about prior activities across models and tools. Agent-to-Agent Protocol empowers autonomous AI agents to coordinate, negotiate and complete tasks directly with one another. Computer use agents allow AI to control user interfaces to complete tasks like executing purchases or filling out online forms.

The duration of tasks that large language models can reliably complete with at least a 50 percent success rate has been doubling every seven months since 2019, according to METR. In 2019, leading models could only manage tasks requiring a few seconds of human effort. In 2025, Anthropic's Claude 3.7 Sonnet boasts a "time horizon" of 59 minutes, capable of completing tasks that would take a skilled human nearly an hour. Claude 4.5 has extended this to more than 30 human hours.

The shift toward AI-mediated commerce could reshape advertising strategies across the industry. Advertising technology expert Karsten Weide suggested that "direct response advertising will fade. Brand advertising will gain in importance as we want to influence consumers before they tell their agent what to do," according to analysis cited by McKinsey. This transformation would prioritize brand awareness campaigns over direct response formats currently dominating digital marketing.

Research published in September 2025 found that most merchants welcome AI-mediated shopping experiences while maintaining control over customer interactions. Amazon represents the most significant exception to merchant openness, explicitly blocking AI bots through robots.txt configurations while developing proprietary alternatives.

Eric Seufert from Mobile Dev Memo argued in August 2025 that blocking AI agents serves Amazon's business interests because agentic commerce "violates the motivations of retail outlets to control the customer relationship and monetize their first-party data with advertising." Amazon and Shopify collectively control more than 50 percent of the US ecommerce market and currently block AI agents to maintain discovery ownership and protect retail media businesses racing toward $300 billion by 2030.

Amazon filed a federal lawsuit against Perplexity AI on November 4, 2025, alleging the startup deployed covert artificial intelligence agents into the e-commerce platform's systems without authorization. The complaint marks the first time Amazon has taken legal action against an AI company over autonomous shopping agents.

The McKinsey report emphasizes that trust transcends mere consumer sentiment and becomes foundational infrastructure in agentic commerce. When an AI agent shops on behalf of a consumer, trust becomes abstract, filtered through layers of data, automation and institutional frameworks. For many consumers in countries like Germany and Japan, where traditional payment methods still dominate, agentic transactions might indeed be a leap of faith too far.

In Germany, account-to-account transfers account for 26 percent of e-commerce payments, alongside a notable 20 percent via buy now, pay later options. This shows a clear preference for methods that emphasize control and transparency. If consumers are already hesitant to share their banking details with a static website or even a familiar payment platform, adoption patterns may vary significantly by market.

The PSE Consulting study was commissioned by Sky Parlour and conducted by market research company OnePoll in accordance with the Market Research Society's code of conduct. Data was collected between November 13 and November 17, 2025. All participants are double-opted in to take part in research and are paid depending on the length and complexity of the survey.

Jones stated that as agentic AI becomes more embedded in everyday shopping, usage will continue rising but emphasized that widespread trust will only come when consumers feel protected and in control. "This has major implications for fintechs, merchants, and payment processors, especially around real-time authorisations, fraud detection, and liability management," Jones stated.

The McKinsey report indicates that businesses must be willing to disrupt their own processes and models to stay ahead. This means rethinking traditional e-commerce strategies and revenue streams to embrace AI solutions that can improve product discovery, customer service and post-purchase experiences. By integrating agentic AI early and investing in the right infrastructure—whether through AI-powered recommendation engines, chatbots for real-time assistance, or smart shopping assistants—businesses can gain a competitive edge by offering an increasingly frictionless, intuitive shopping journey.

The shift toward agentic commerce is likely to happen quite a bit faster than the e-commerce shift. In 1999, 100 million internet users (just under 2 percent of the global population) began exploring e-commerce. In 2025, nearly 5.6 billion users are connected to the internet, representing approximately 68 percent of the world's population. In this hyperconnected age, dramatic shifts at a much larger scale can occur much more rapidly.

Timeline

• May 2025: Cloudflare shares Web Bot Auth proposal for agent authentication
• August 30, 2025: Amazon blocks AI bots from major tech companies including OpenAI, Anthropic, Meta
• September 16, 2025: PSE Consulting research reveals major merchants welcome AI agents while Amazon blocks competition
• September 17, 2025: Amazon introduces agentic AI across seller platform transforming Seller Assistant
• September 29, 2025: OpenAI launches instant checkout for ChatGPT with Stripe partnership and Agentic Commerce Protocol
• October 6, 2025: Independent analyst questions agentic commerce viability despite ChatGPT checkout launch
• October 15, 2025: Six companies launch Ad Context Protocol for agentic advertising
• October 24, 2025: Cloudflare partners with Visa and Mastercard to develop security protocols for AI agents
• November 4, 2025: Amazon demands Perplexity stop AI shopping in Comet browser and files federal lawsuit
• November 13, 2025: IAB Tech Lab opens Agentic RTB Framework for public comment
• November 18, 2025: Amazon deploys generative and agentic AI across shopping platform with 250 million Rufus users
• November 25, 2025: PSE Consulting releases UK research showing 49 percent of adults use AI regularly and 22 percent plan AI-assisted holiday shopping
• November 2025: McKinsey releases comprehensive report projecting $3-5 trillion global agentic commerce opportunity by 2030

Summary

Who: PSE Consulting, a leading payments consultancy, commissioned research of 2,000 UK online shoppers conducted by OnePoll. McKinsey released parallel analysis examining the global agentic commerce landscape. Chris Jones, Managing Director at PSE Consulting with over 20 years of payments industry experience, provided commentary. Major technology companies including OpenAI, Stripe, Google, Visa, Mastercard and Cloudflare are developing infrastructure for autonomous AI shopping.

What: Research reveals 49 percent of UK adults now use AI tools regularly, with 22 percent planning to use them for Black Friday and Christmas shopping. Among those planning AI-assisted shopping, 85 percent would trust AI to place orders and execute payments autonomously. McKinsey projects agentic commerce could orchestrate $900 billion to $1 trillion in US B2C retail revenue by 2030, with global projections reaching $3 trillion to $5 trillion. Major payment networks and technology platforms have launched authentication protocols, payment frameworks and security standards to enable AI agent transactions.

When: The PSE Consulting study was conducted between November 13 and November 17, 2025, with results released November 25, 2025. The McKinsey report was published in November 2025. OpenAI and Stripe launched the Agentic Commerce Protocol on September 29, 2025. Google introduced Agent Payments Protocol in September 2025. Cloudflare announced collaboration with Visa and Mastercard on October 24, 2025. The Agentic Commerce Protocol is expected to reach Europe within six to nine months from late 2025 due to Strong Customer Authentication requirements.

Where: The PSE Consulting research focused on UK consumers, showing particularly strong adoption among 18-34-year-olds in urban areas. McKinsey's analysis encompasses global markets, with specific projections for the US B2C retail market. OpenAI's instant checkout functionality currently operates in the United States for both buyers and sellers in its initial phase. Payment infrastructure developments span multiple jurisdictions, with different adoption patterns expected in markets like Germany and Japan where traditional payment methods still dominate.

Why: The convergence of mature AI capabilities, extensive internet connectivity (5.6 billion users representing 68 percent of global population), and standardized payment protocols enables AI agents to execute complex shopping tasks autonomously. Systems designed for human-paced transactions now face pressure to support high-frequency, autonomous agent-initiated flows, necessitating fundamental infrastructure changes. Privacy concerns (49 percent), fraud risks (46 percent) and accuracy worries (41 percent) remain significant barriers, but 85 percent trust among early adopters signals latent demand for autonomous commerce. The shift matters for marketing professionals as it transforms customer relationships, advertising strategies and revenue models across retail, payments and advertising technology sectors.