Colin Baines from Worldpay warns Australian businesses must prepare for agentic commerce, where AI agents autonomously make purchasing decisions. Friction at checkout could mean lost sales to algorithms.
What’s happening: Colin Baines, Country Manager and Vice President of Commercial for Australia and New Zealand at Worldpay, warns Australian businesses must prepare for agentic commerce, where AI-powered agents autonomously discover, evaluate and execute purchases on behalf of consumers without direct human input.
Why this matters: AI agents prioritise speed, reliability and measurable performance over traditional marketing persuasion. Businesses that fail to optimise payment systems, eliminate checkout friction and ensure transparency risk becoming invisible to the algorithms increasingly making purchasing decisions, regardless of product quality.
Consumer behaviour has never been static, according to Colin Baines. From shopping malls to online websites and now social media marketplaces, each shift has redefined how people discover and buy.
Now, Baines argues, we are entering a new phase of commerce: agentic commerce. AI-powered agents are now able to act on behalf of consumers to make shopping decisions, from comparing and selecting to, in some cases, completing purchases automatically, all without direct human input.
For Australian businesses, this means there is a new playing field. Baines warns that AI agents are changing the rules of the game and transactions now demand speed, reliability and seamless processes. Merchants must ensure their systems can meet these requirements to remain competitive.
Autonomous decision-makers
AI agents are not simply digital assistants, Baines explains. They are autonomous decision-makers, and that is changing the way we approach e-commerce and shopping in general. AI agents are increasingly responsible for discovering, evaluating and executing purchases on behalf of consumers.
Discovery is shifting away from search engines and advertising but instead relying on generative AI interfaces that understand user preferences. Decisions are guided by structured data rather than human impulse, with agents weighing price, convenience, delivery speed, reliability and even ethical considerations before completing a purchase.
Trust is also being redefined. Consumers now expect transparency, secure payment flows and visible reliability. Marketing messages or persuasive copy that might have influenced a human buyer hold little sway over AI agents, according to Baines. These systems prioritise measurable performance signals and consistency.
“This signals a decisive shift and businesses that recognise this transformation and adapt their processes now will gain a clear advantage, while those that rely on traditional approaches risk being overlooked, not because of product quality, but because their operations are invisible or unreliable to the very agents making buying decisions,” Baines writes.
Performance determines success
Friction at checkout has long caused human shoppers to abandon purchases. According to Worldpay’s Retail Payments Performance Report, more than half (54%) of shoppers give up if they have to retry the payment. As AI agents begin to take a more active role in shopping, Baines suggests the tolerance for error is likely to be even lower. Any delay, unsupported payment method or security gap could mean the transaction is dropped automatically.
This makes optimised payment systems non-negotiable, he argues. Tokenisation is no longer optional; it is the foundation of agentic commerce. By replacing sensitive card details with secure tokens, businesses allow AI agents (and humans) to transact safely, whether for one-off purchases, subscription renewals or recurring bills. Consumers also rely on visible trust cues such as familiar payment brands, multiple checkout options and biometric authentication, all helping reinforce confidence in the purchase process.
At the same time, liability and fraud add another layer of complexity. Traditional chargeback frameworks were designed with human buyers in mind and are not fully equipped to handle autonomous agents that may misinterpret preferences or exceed budgets. Fraud detection, likewise, must evolve alongside these new behaviours.
What may once have looked like suspicious bot activity may now be a legitimate AI agent completing a purchase on behalf of a consumer. Next-generation fraud systems need to monitor the full customer journey and distinguish legitimate agent transactions from malicious activity. Businesses that are able to offer such intelligence will maintain both trust and operational security, Baines notes.
“The stakes are high. As ever, consumers increasingly expect speed, reliability, and transparency but with the rise of AI agents in the purchase process, businesses that fail to meet these expectations face an even bigger risk of losing sales than before,” he writes. “Every technical upgrade, every optimisation of checkout flow and every reinforcement of security is a metric that could sway an agent from one merchant to another.”
New opportunities emerging
While these changes are significant, they also create opportunities, Baines argues. Consumers are most likely to delegate recurring and low-risk purchases, such as subscription services, billing, ticketing and delivery, to AI agents first. Businesses that operate in these spaces and make these experiences frictionless and reliable will have an immediate opportunity to lead, provided that their platforms are AI agent-ready.
Beyond these transactions, any merchant that builds systems prioritising speed, reliability, transparency and security can gain a competitive edge. In a world where persuasion is replaced by performance, eliminating friction is the fastest path to visibility and customer loyalty, according to Baines. Streamlined checkout flows, clear security cues and flexible payment options signal to both consumers and AI agents that a business is dependable, which can directly influence purchase decisions.
The shift towards AI-driven commerce reflects broader trends in Australian retail. Recent research shows that AI personalisation now outperforms traditional loyalty programmes by nearly 15 percentage points in Australia, with 85% of consumer goods leaders believing AI agents will be essential for competition within two years.
Algorithm-governed marketplace
Businesses that adopt these practices early not only capture repeat customers but also position themselves as trusted partners in the AI agent ecosystem, Baines argues.
“Agentic commerce isn’t coming, it is already here and businesses that invest in tokenisation, intelligent fraud detection, and flexible payment infrastructure will be prepared to thrive in a marketplace increasingly governed by algorithms,” he writes. “Those who wait, risk being invisible to the very agents that will increasingly make purchasing decisions on behalf of their customers.”
The question is no longer whether AI agents will reshape commerce, according to Baines. “The real question is whether your business will stand out when the algorithms start making selections.”
Keep up to date with our stories on LinkedIn, Twitter, Facebook and Instagram.
