The rapid expansion of the accounting technology market presents practice leaders with a complex challenge. While the operational benefits of automation and AI are clear, the sheer volume of choice creates a high-stakes decision environment. With new AI tools launching every week practice leaders must sift through these options to identify solutions that rigorously protect client confidentiality and adhere to complex accounting laws.
In our recent Leading Voice broadcast, Stuart Miller (Director of Public Policy and Technology Research at Xero) posed a critical question regarding this selection process. He asked how firms can choose a tech stack with “enough confidence” to ensure the tools they migrate clients to “won’t disappear in the next sort of 3 to 6 months”.
The trap of “free” technology
Alex Falcon Huerta, founder of Soaring Falcon, identifies the allure of cost-free software as a primary risk factor for modern firms.
She notes that while “it’s easy to go” because of the low barrier to entry, this approach often necessitates a dangerous oversight regarding data privacy. When a tool is free, the user is often the product, or their data is being used to train the model.
Alex explicitly warns practice leaders to “stay clear of those ones,” emphasizing that the primary consideration must always be the data protection of our clients.
The due diligence checklist
To navigate this risk, Alex advocates for a rigorous selection process that prioritizes stability over novelty. She advises looking for tools that are already established and backed by a trusted company rather than ephemeral startups that may not survive market consolidation.
Beyond corporate stability, the non-negotiable criteria must be regulatory compliance. Any AI tool introduced into a practice must align with all of the accounting laws alongside AML laws and GDPR rules and regulations.
If a tool cannot prove it adheres to these strict standards, it simply cannot be part of the professional tech stack. The risk of a data breach involves more than just financial cost; as Alex noted earlier in the broadcast, the “brand and reputation for that firm will be taken through the mud” if client data is lost or mishandled.
The “beta test” protocol
Even when a tool appears compliant, Alex recommends a tiered testing protocol before rolling it out to the wider firm or client base.
Her approach involves being the first subject of the experiment. She will always test it on herself before asking trusted clients if she can beta test it with them. This ensures that any workflow issues or data quirks are identified in a controlled environment.
This aligns with Stuart Miller’s perspective on “responsible data use commitments.” He argues that AI use within a firm must be gated and private, ensuring that open-source models aren’t utilizing all our own chats and information to push things through.
Building for long-term integrity
Innovation in accounting requires a secure foundation. While the pressure to adopt AI is high, the responsibility to protect client data is higher.
By ignoring the temptation of free tools and adhering to a strict framework of GDPR and AML compliance, practice leaders can build a tech stack that delivers efficiency without compromising the firm’s integrity. As Stuart Miller summarized, we cannot take client data lightly; it is worth the effort to “make sure we get it right.”