Byte-Sized AI is a bi-weekly column that covers all things artificial intelligence—from startup funding, to newly inked partnerships, to just-launched, AI-powered capabilities from major retailers, software providers and supply chain players.
Mastercard, Visa unveil plans to partner with tech giants on AI shopping
Two of the world’s largest payment providers, Mastercard and Visa, announced this week that they will work with some of AI’s largest companies to able agentic shopping experiences for consumers. Agentic shopping will happen when a consumer tells a model to buy an item for them, and the model obliges; for instance, if a consumer wants to purchase a black T-shirt made of 100 percent cotton, an AI agent can help find suitable options, and, in the future, will be likely to purchase directly for the consumer.
Both companies are tokenizing consumers’ physical payment information to make that happen—that is to say, they’re creating digital tokens that can be used by AI agents to make secure purchases without giving up sensitive consumer information.
Jack Forestell, Visa’s chief product and strategy officer, said that kind of consumer protection will enable better outcomes for agentic shopping, while also garnering consumer trust.
“Soon people will have AI agents browse, select, purchase and manage on their behalf,” Forestell said in a statement. “These agents will need to be trusted with payments, not only by users, but by banks and sellers as well.”
Visa announced it has partnered with Anthropic, IBM, Microsoft, Mistral AI, OpenAI, and Perplexity, while Microsoft is starting with a Microsoft partnership it plans to expand to other providers in the future.
The providers made it clear that consumers will be able to control how their money is spent; per Visa, consumers will be able to “set spending limits and conditions, providing clear guidelines for agent transactions.” And, according to Mastercard, users will retain “complete control over what the agent is allowed to purchase on their behalf, ensuring that the payments they make are securely authorized and identified.”
While the technology is likely to have myriad use cases once agentic AI becomes more prevalent in shopping, Mastercard knows it can be used for simple consumer transactions and for B2B use cases. In its announcement, it shared an example about a small textile business.
“A small textile enterprise will be able to use their AI agent to handle sourcing, optimize payment terms and manage logistics with an international supplier. From there, the AI agent can complete the cross-border purchase using a Mastercard virtual corporate card token and arrange for cost-effective, expedited delivery,” Mastercard noted.
The companies’ announcements come in the same week OpenAI announced consumers would have access to refined shopping features meant to simplify product discovery and purchasing journeys for e-commerce.
Skechers launches AI-powered retail assistant in Singapore store
Skechers has teamed up with We Are Social Singapore to bring an AI-powered assistant into a physical store in Singapore. The assistant, which it calls Luna, has the ability to take into account customers’ preferences—alongside what they are actively wearing—to recommend products, and how to style them, for that consumer.
Luna interacts with customers through a kiosk or through messaging app Telegram, and can point consumers in the direction of in-store products or online items.
Manolis Perrakis, innovation director at We Are Social Singapore, said the activation is a display of the way AI agents may start interacting with consumers in real time, both physically and digitally.
“The emergence of AI speech-to-speech technologies is powering an agentic AI revolution that forms the backbone of future consumer-facing systems. Luna is an additional touchpoint for Skechers to complement its innovative retail experience, uniting retail and online environments into a seamless ecosystem. This integration empowers brands to strengthen customer relationships and drive dynamic, two-way conversations that connect the shop floor and customers’ phones.”
The activation is part of a short-term event to celebrate the opening of the store within the Punggol Digital District, so Luna will be a temporary fixture. Nonetheless, the assistant is another way brands have begun using AI-powered technology to interact directly with their consumers, though Luna is more novel than, say, using generative AI to power customer service chatbots for digital.
Skechers latest experiment is evidence that brands continue to refine their approach to consumer personalization, particularly in a competitive retail environment. But as brands do so, they have to toe the line between helpful and creepy; data from Boston Consulting Group shows that two-thirds of consumers that have experienced AI-driven personalization said their encounters felt invasive or inaccurate.
Phia wants to give consumers accessible ways to shop secondhand
Sophia Kianni and Phoebe Gates announced late last month that they had launched Phia, which uses AI to give consumers insights on whether there’s a way to buy a product they’ve viewed at a lower price.
Users can download the Phia app or use the technology through extensions on desktop. When viewing an item on a brand or retailer’s site, consumers using Phia are presented with a button that asks, “Should I buy this?”
From there, Phia generates a price summary—whether the price is low, typical or high—and suggests where a consumer can buy an identical or similar product for cheaper. It shows both options to buy new from sites like Amazon and to buy secondhand from marketplaces like Poshmark, Depop and eBay.
The founders noted that their proprietary technology works for any e-commerce brand.
Kianni said Phia helps encourage users to buy secondhand items; by showing them a variety of resale items, the technology wants to improve environmental outcomes and help consumers, who currently lack optimism about the U.S. economy, pay less for the goods they want or need to purchase.
“We, like so many consumers, want to shop smarter and make the most of our money,” Kianni said in a statement. “Great secondhand options exist, but they’re scattered across hundreds of websites—and no one has time to search them all. Our patented model solves that pain point by delivering instant price insights and better options to help customers make smarter, faster decisions with their money.”
Oracle uses AI to quell logistics’ leaders trade concerns
As companies fret over how their supply chain practices will be impacted by tariffs and economic uncertainty, technology providers want to use AI and automation to preemptively stop their headaches.
Oracle announced in late April that it had added trade management capabilities to its supply chain and manufacturing cloud, focused on ensuring clients can adequately handle the changes to their operations onslaught by tariffs and trade agreements.
It does so by automating certain processes, like product classification—the technology can help logistics managers easily understand which codes to use when importing new or altered products. It also pulls together information on foreign trade zone status, tracks imports or exports to create insights about duty drawback and generates reports to file such paperwork later on.
Chris Leone, executive vice president of applications development for Oracle, said the new tool can help ensure leaders don’t have to keep up with every trade update manually.
“Supply chain leaders are rising to the moment by seeking new ways to manage their business with global trade agreements and international tariffs in a state of flux. To help our customers with this complexity, we have added new capabilities within Oracle Global Trade Management that enable supply chain leaders to quickly respond to changes and minimize disruption to their global supply chains,” Leone said in a statement.
