Since Gartner's Don Scheibenreif and Mark Raskino coined the term "machine customers" the ability for consumers to outsource purchasing decisions has been a slow boil.
Machine to machine purchasing has existed in B2B for some time, through RosettaNet and other technologies. In the B2C world, auto-replenish brought subscription economics to millions of consumable items, even if it did become one of the biggest reasons for refund requests. But true machine buyers, have taken longer to arrive.
Machine customers, algorithmic buyers, agentic commerce and AI-first journeys all build on the same idea – the use of AI to select and buy products with varying levels of autonomy. While it's a development that directly and fundamentally changes CX, it isn't within CX's control. Instead, it is an innovation developed and delivered by big tech; the likes of Anthropic, Google, OpenAI, Meta, and others.
And it's developing rapidly.
When ChatGPT 3.5 launched in 2022, millions adopted the tool and started to use it to do everything from write code, to draft emails to customer service departments.
Now in mid-2026, we have AI assistants and LLMs that can search for and shortlist products then drive traffic to retailer websites, we have the ability to buy within some chat interfaces, and we even have payment protocols.
Now, Meta and Google are testing AI agents for consumers on an even bigger scale than OpenClaw. Once launched it means consumers who are not necessarily savvy with technology will have the ability to instruct AI agents to execute tasks on their behalf – including buying products and services. The challenge is no longer the tech; it's building in the security and backstops that underpin customer trust.
This article explores what the development means for agentic commerce and CX, and the key actions organizations need to take, while there is still time to take them.
How AI agents for consumers will change the game for CX
In early May 2026, it was reported both Meta and Google were building AI agents for non-business users. Both organizations said their agents will be able to conduct daily tasks on behalf of users and, while the use cases will span everything from summarizing the news to following stocks, there will be obvious implications for commerce and, therefore, CX.
The big one is machine customers.
Consumers have long cherished convenience in CX, and early expectations for machine customers focused on assistants that could handle the "boring stuff", account updates, service interactions, and complaint emails. Autonomous buying was a more distant dream. Now, however, it's closer to becoming a reality.
"We're likely to see consumer AI move from 'help me decide' to 'just do it for me', where people increasingly interact with AI instead of websites or apps," says Simon Thorpe, director of global product marketing - customer service and sales automation at Pegasystems. "That means AI isn't just answering questions anymore - it's taking on the discovery, making recommendations, and even completing transactions like buying, booking, or paying."
While this is a major leap for consumer tech, it also comes with huge implications for CX.
"As a result, the battleground shifts away from front-end experiences to back-end execution," Thorpe explains. "Who can actually deliver outcomes safely, accurately, and fast? And that's where trust becomes critical - if AI agents are acting on a customer's behalf, you need clear authorisation, validation, and accountability built in, not just a great UX. Ultimately, the winners won't be the ones with the best chat interface, but the ones who can actually orchestrate the work, policy, and execution in a way that's reliable, transparent, and trusted at scale."
As outlined in this CX Network article, having a track record for CX excellence is now more important than a flashy marketing campaign.
In preparation, Thorpe says organizations need to "rethink journeys as executable outcomes".
"Because when AI agents are driving more interactions at scale, you'll more than likely see more volume flowing through and more edge cases to handle," he adds.
This means it's no longer about guiding a customer step-by-step through a journey, but about "reliably executing the outcome they're trying to achieve, whether that's buying, changing, or resolving something", Thorpe says.
"That also means treating AI agents as a new channel in their own right and handling agent-initiated requests alongside human interactions in real time," he adds.
This CX Network webinar explains more about how agents are now the channel.
Bringing machine customers to life
Although the availability of true machine customers remains in the hands of the major tech players, there are things organizations need to do to ensure they are not caught off guard by algorithmic buyers and autonomous decision agents working on behalf of customers.
First of all, as outlined in CX Horizons: The state of CX in 2026, brands need to prepare for four interaction modes.
"Brands can't ignore this," says Sue Duris, principal consultant at M4 Communications. "They must prepare and enable four interaction modes: human-to-human, human-to-machine, machine-to-human, and machine-to-machine and be ready to switch from one to another."
Organizations also need to ensure they put the right foundations in place around identity, control, and trust.
Thorpe says: "You need to be able to verify who the agent is and who it represents, define clear permissions around what it can do and how much it can spend, and connect into interoperable commerce and payment rails so it can actually act across ecosystems.
He adds: "At the same time, everything has to be regulation-ready, with built-in transparency, oversight, and auditability to support safe adoption at scale."
Underpinning all of this is the need for an enterprise orchestration layer that can coordinate decisions, workflows, and execution behind the scenes. Thorpe adds: "The organizations that succeed will be the ones that can prove consent, enforce limits, and govern end-to-end accountability with confidence."
The potential for machine customers has existed for some time, but the tech to make it a reality was still underdeveloped. Now, with two of the biggest tech players in the world actively testing solutions, CX is on the verge of sweeping change – and few are prepared.