While small business solutions company
"2026 is about moving from AI as a feature to AI as the core engine. We aren't just shipping tech for the sake of it; we're focused on driving adoption by making these tools (from auto bank reconciliation to real-time financial insights) a seamless, essential part of the daily workflow," she said.
Jolly framed this not as a departure but a continuation of the company's ongoing technology strategy. Overall, she said, the AI strategy rests on four pillars: automated actions and workflows, actionable insights, reimagined experiences, and trust. To these ends the company has been developing AI agents to automate routine tasks across workflows and provide personal intelligence for things like cash flow scenarios, as well as emphasize the AI agent-management aspects of the SaaS experience, which includes leaning heavily on security, privacy and expertise. This is all in the service of evolving Xero from a system of record to something that enables actionable decisions.
"We'll also continue to broaden and deepen these capabilities, continuing to evolve Xero into a system of action and decision-making," said Jolly. "By automating more of their key jobs to be done, we're saving customers more time and providing intelligent tools they need to manage and grow their business with greater confidence."
That confidence is important. She noted that the technology has advanced rapidly in a short time, which means more people than ever are using AI. Even within Xero, she said, the use of its JAX AI, which was first
"While the technology is moving at an incredible pace, recent developments haven't forced a pivot; they've actually validated our direction," said Jolly. "In an era of generic AI hype, trust is the ultimate differentiator. Small business accounting isn't a place where you can 'hallucinate' or guess. It's built on what we call decisioned data, a series of high-stakes decisions that build on each other, from reconciling a transaction to closing your books and filing taxes. To get this right, you need layers of data connected by a context graph that understands the ledger line-by-line. This is what allows our AI agents to understand exactly how data should be used to automate a task safely and accurately."
She said this is why Xero has focused on what she called "accountable intelligence." AI, according to Jolly, needs to show its work behind every answer at every step of the way, or else customers will not have the certainty needed to reliably use it. She said Xero is "doubling down" on this approach as a result. This is especially pertinent as Xero moves its automation capabilities past mundane repeatable tasks and into more complex workflows.
"As we move into more complex workflows, our approach is centered on being accountable. To us, accountable intelligence means the AI handles the heavy lifting and surfaces the insights, but the person remains the strategic architect who provides the final context. Because Xero is a trusted system of record with built-in audit trails and safeguards, our AI is grounded in truth, not guesswork," she said.
Part of this comes from eschewing general AI models and focusing instead on specialized applications derived from the architecture of the system itself.
"You can't just throw a generic AI tool at accounting; the stakes are too high, and the compliance world is just too complex," said Jolly.
Another part comes from cooperation with local CPA firms. While Xero allows people to do much on their own, the advisory abilities of human professionals can play a large role in ensuring accountability and trust in their AI systems.
"AI is a massive lever for local CPA firms. Our strategy is to ensure they're armed with an approach focused on accountable intelligence—where the AI handles the routine heavy lifting, but the advisor provides the essential human context. This will be especially critical for firms facing the industry's persistent talent shortage, as AI increases their capacity to serve clients and boost profit margins, without increasing overhead," she said.
The need to ensure safety and reliability in their AI systems is further underscored by what Jolly said would be an increasing focus on payments this year after
"It's no longer just about recording a bill. Just Ask Xero (JAX) can now assist with the whole process, from capturing the data to making the payment and automatically reconciling the transaction. This is the system of action in its most practical form," she said. "Because we're talking about moving money, trust is everything. This is a perfect example of how we approach accountable intelligence. We are stripping away the administrative friction that causes late payments and cash flow anxiety, but we're doing it with the safeguards and transparency required to keep the business owner and their advisor in the driver's seat. We aren't just making the process faster; we're making it more certain."
Asked about how the strategy varies by market, Jolly noted that while the overall strategy is global, the context is always local. In the U.S. specifically, she said, that context is strongly defined by an evolving CAS ecosystem enabled by automation advances that allow accountants to focus on higher-level advisory work.
"We often say that while AI is increasingly good at the 'what,' like processing transactions and spotting patterns, it can't replicate the strategic 'so what?' A dashboard can flag a cash flow gap, but it doesn't understand a client's specific risk tolerance or their family's long-term goals," said Jolly. "Only a trusted advisor can provide that lens. By evolving Xero from a system of record to a system of action, we aren't just giving U.S. advisors better tools; we're giving them the data they need to be the strategic architects of their clients' futures."
