Agent artificial intelligence is embedded in everyday shopping and enterprise workflows, turning simple prompts into fully executed tasks and streamlining online interactions that improve customer-to-supply chain interactions.
The promise to beckon is handoff commerce, whether you are a consumer or a business, but reality remains mixed. At B2C, retailers deploy AI agents across customer interactions. In B2B, adoption is measured to help businesses weigh security and governance. Agent AI is no longer just a promise. It is a progressive reality, but it is still carefully shaped.
Walmart''Super AgentFor example, now strategy customer, supplier, fellow and Developer Interactionaims to make AI the main interface between commercial and operation.
“Agents help us automate and simplify almost everything we do,” says Walmart Chief Technology Officer. Suresh Kumar With the July event, adding timing, almost everything uses AI to reflect “how customers are ready.”
SparkyRetailer Customer Assistant handles tasks from product recommendations to budget planning.
Second quarter 2025 Revenue In August 2024, Walmart revealed that Generator AI can help enrich or improve over 850 million catalog data points. This is a task that requires 100 times more AI-free workforce.
Thursday (August 21), the company's second quarter, 2026 Revenue Please call the CEO Doug McMillion Sparky said it will develop agent functionality over time.
“As we improve and expand Sparky, we'll become even smarter and more personalized,” he said. “It will be the main digital vehicle for managing everything from discovery, shopping and reload to return. The other superagents we're building include those for associates who bring everything from scheduling to sales data into one place. [There’s also] One is for suppliers, sellers and advertisers that you use to manage onboarding, orders, campaigns, and more. ”
rival Amazon Agent AI is also moving forward. The company said Agent AI Autonomous software agents acting on behalf of customers or users are key drivers of future growth. These AI agents are treated as enterprise customers using aws A tool for automating retail transactions.
Amazon CEO Andy Jussy In its second quarter revenue call on July 31, the company said it was preparing enterprise AI agents to shape demand across its platform.
B2B conversion via agent AI
in nod For shifts occurring in commercial settings, Transcard Integration Agent AI Joined the Smart Exchange platform in April to automate vendor onboarding, Know your business (KYB) Process. AI digitizes vendor management with high-touch, customized interactions while reducing manual steps.
In another example, Infosys BPM He said it was unfolded in May. AI Agent Among them Accounts paid in the cloud Solution, that It's driven By doing so Topaz Generation AI and azure's ai stack. This solution autonomously handles invoices with more accuracy and efficiency.
Mastercard I launched it Agent Payment May initiative, bringing agent payments to the forefront. This offering introduces an agent token that allows AI agents to trade on behalf of trusted and secure users. MasterCard is affiliated with Microsoft, IBM Others expanding Agent Commerce into enterprise operations.
visa's Intelligent Commerce API Embed agent functionality directly into transaction flows, allowing AI agents to autonomously shop and pay.
in B2B,The deployment of autonomous AI agents poses complex security and governance challenges that extend beyond traditional automation. Risk, trust and monitoring are important, especially in high stakes areas. Like Finance and procurement.
The B2B ecosystem is “built on deep trust and low risk tolerance.” Autonomous pharmacology systems need to be controlled carefully Pymnts wrote on June 10th to avoid catastrophes such as incorrect payments and contract breakdowns.
Reliable Chief and Compliance Officer Kathryn McCall Shift:Governance revolutionencouraging the adoption of “bound autonomy” and includes sandbox environments, scope permissions, audit logs, human loop controls, and forensic replays.
AI agents should be treated as nonhuman actors with unique system identities with explanability, surveillance and kill switches, she said.
In the retail and business as a whole, Agent AI has evolved from a proof of concept to a basic infrastructure. As ai It will evolveit will become increasingly defining commercial transactions It will startit's running and finished.
Subscribe daily for all PYMNTS AI coverage AI Newsletter.
