Australian Payments Plus (AP+), the organization that runs much of Australia's payments and identity infrastructure, has shared the results of rolling out OpenAI's ChatGPT Enterprise and its AI coding tool Codex across its work[1]. In a highly regulated financial-infrastructure environment, how far can AI push the speed of investigation and document work? This article walks through the concrete use cases alongside the numbers from the company's internal survey.

What Australian Payments Plus Does

Australian Payments Plus is a core organization behind Australia's payments and identity systems. It operates the eftpos debit network, BPAY bill payments, the NPP (New Payments Platform) real-time payments infrastructure, and the ConnectID identity verification service, connecting more than 150 organizations spanning banks, retailers, government agencies, and fintechs[2]. It was formed in 2022 through the amalgamation of eftpos, BPAY, and NPP into a single member-owned entity[2].

Because of the nature of this work, AP+ employees handle a huge volume of information across scheme rules, technical specifications, member obligations, operational processes, cybersecurity, and regulatory expectations. Speed matters, but accuracy and accountability matter even more[1]. To move through that complexity faster, AP+ rolled out ChatGPT Enterprise company-wide and then extended Codex into product, engineering, and technical workflows[1].

Faster Information Discovery and Drafting with ChatGPT Enterprise

Before the rollout, answering a question or moving work forward often meant manually searching across scheme rules, technical specifications, and internal documents. In payments, a small detail can change the answer, so this information hunt was no small task[1].

Today, AP+ teams use ChatGPT Enterprise to summarize complex material, draft data-driven member communications, and structure ambiguous problems. For example, by navigating eftpos specifications and related documents with ChatGPT, employees can find the right starting point faster before applying expert review[1].

Meeting notes become summaries, workshop inputs become decision materials, and dense design documents become solution overviews. AP+ frames ChatGPT not as a way to automate work but as a sparring partner for sharpening thinking[1].

The numbers reflect this. Of the employees using ChatGPT, 77% saved 2 or more hours a week, and 80% reported improved creativity or work quality[1]. Internally, staff have created more than 300 custom GPTs and more than 1,000 Projects, pointing to broad adoption[1].

Shorter Investigations and Prototypes with Codex

On the technical side, the Codex coding tool is being used for complex investigations. In one reconciliation case, AP+ teams used Codex to trace a subtle timestamp inconsistency across system logs and reconciliation data, reducing what had been days of manual investigation to minutes[1]. Investigation time for complex reconciliation issues fell from about 4 hours to roughly 30 minutes[1].

Product development is changing too. Earlier testing tended to rely on click-through screens that showed only what an experience might look like, not how it would behave. Now, by combining ChatGPT Enterprise and Codex, teams can simulate payment journeys, mobile interactions, authentication flows, and checkout experiences in environments that behave closer to real systems[1]. Building a working simulation now takes 1 day, down from what could previously take days to weeks[1].

AP+ is also exploring whether Codex can support security teams in areas such as threat modeling, vulnerability analysis, alert triage, and visibility across interconnected systems, always keeping human experts accountable for risk decisions, validation, and response[1].

Lessons from Adopting AI in a Regulated Industry

AP+ offers practical lessons for regulated organizations looking to scale AI responsibly. The company highlights four points[1].

First, make the secure path the easy path: give employees governed tools with appropriate access so they can experiment safely within clear boundaries. Second, make governance a launch partner: in a regulated environment, adoption depends on high-quality conversations with the teams accountable for privacy, security, governance, and operational risk. Third, let teams learn in context: adoption worked best when employees saw relevant examples from their own teams rather than generic training alone. Fourth, use champions to make change tangible: bringing use cases into existing team rhythms reduced the need for extra meetings or standalone AI programs.

As ChatGPT Enterprise becomes part of daily work and Codex adoption grows, AP+ is expanding how it uses AI across product development, technical investigation, and member-facing workflows, always with expert review and human accountability at the center[1].

Summary

In payments, where accuracy comes first, Australian Payments Plus applied ChatGPT Enterprise and Codex to information discovery, drafting, complex investigation, and prototyping, reporting outcomes such as saving 2-plus hours a week and sharply cutting investigation time. Framing AI as a tool that supports expert judgment rather than replacing it makes this a useful reference case for adoption in regulated industries.

Source: https://openai.com/index/australian-payments-plus

Source: https://www.auspayplus.com.au/