The ethical technologist? The training imperative for public finance in the age of AI

Artificial intelligence (AI) has the potential to enhance tax administration, improve expenditure forecasting and streamline public procurement, but its deployment must be guided by strong regulatory frameworks and ethical oversight.

ASEAN countries will need to carefully navigate the integration of AI into their public finance functions - by balancing opportunities for efficiency and transparency with significant risks around governance, data security, and algorithmic bias.

The rapid pace of AI development makes it important to seek out best practices from around the world.  

According to a new report into AI adoption by the accountancy body ACCA, public auditors in the UK are currently tapping several generative artificial intelligence (GenAI)-powered tools to improve the effectiveness and efficiency of their audit management.

Developed by the Data Analytics team at the UK’s Government Internal Audit Agency (GIAA), these tools include Risk Engine, which helps auditors identify risks related to a user’s business objective; as well as Writing Engine, which reduces the auditor’s manual effort to write audit reports.

The best practices around public sector use of GenAI took reference from the UK government’s AI framework.

“Innovation and oversight must go hand in hand to ensure that AI in public finance is both effective and accountable,” said the report’s author, ACCA’s Senior Subject Manager – Technology, Alistair Brisbourne.

Good AI governance and oversight provide clear boundaries and guidance for public sector officials, which can be “a real boon for innovation,” he explained.

He shared more about how public finance professionals can keep up in the age of artificial intelligence (AI) to ensure effective and responsible governance of public funds.

To subscribe to the GovInsider bulletin click here.

Tackling the dual challenge of skills and ethics

As AI systems become increasingly sophisticated and widespread, Brisbourne underlined the need for governments to invest in training and development to ensure that public finance professionals can work effectively alongside emerging technologies.

According to him, training should build on existing strengths.

“AI tools augment our capabilities while raising new questions concerning control, reliability and professional responsibility. These capabilities don't replace professional judgment – they underscore its importance.”

In other words, training needs to improve our ability to apply professional scepticism and critical thinking to question, analyse, interpret and evaluate to make sound judgement calls.

Brisbourne points out that technical proficiency and ethical reasoning should be key considerations to achieve these aims.

Technical proficiency includes AI and data literacy, which enables finance professionals to understand how different AI models function, interpret AI-generated insights, and make informed decisions based on data-driven analyses.

The other area that the training should cover is methodological, which refers to the ability to assess the application of different algorithmic methods and interpret probabilities, he added.

As for ethical reasoning: “Public finance professionals must be able to critically assess the ethical implications of AI-driven financial decisions, ensuring that AI is used responsibly and transparently,” Brisbourne said.

While AI can assist in decision-making, it should never replace the critical thinking, ethical reasoning, and professional scepticism that finance professionals bring to the table.

Ensuring that AI remains a tool for enhancement, rather than a substitute for professional expertise, is essential for maintaining accountability in public finance.

He also emphasised the importance of collaboration between public finance professionals and AI specialists.

In the UK’s audit agency case study, the data analytics team at GIAA worked closely with the auditors to explore how to adapt the GenAI tools into existing workflows and processes.

“Cross-functional teamwork will become increasingly important in designing, implementing, and monitoring AI-driven financial systems,” he noted.

At the same time, we are still in the process of understanding the most valuable opportunities that AI presents, how our organisations need to adapt to integrate new capabilities, and how individual roles and responsibilities will evolve.

The report, The Smart Alliance: Accounting expertise meets machine intelligence, is the latest in a series ACCA has produced to improve understanding for accountants and finance professionals around how organisations are currently using AI, tackling challenges to adoption, as well as explore important topics including how we can maintain trust, upskill, address sustainability challenges, and more.

ACCA’s Policy & Insights research is shared across the world with standard setters, policy makers and regulators.

As the largest global accountancy body serving the public sector, ACCA’s vast membership provides a highly skilled network of that are able to feed into its insights work, giving it the depth and diversity that makes it relevant to people who matter.  

An upcoming issue (released February 28, 2025) explores questions of risk and responsibility.

Private sector involvement in public projects – to what extent?

A balanced approach between building in-house expertise and leveraging external partnerships is key to eventually developing the internal capacity of agencies to understand, implement and manage their own AI-driven financial systems, said Brisbourne.

On private sector involvement in capacity building, engaging external consultants and technology partners can provide valuable expertise, especially for public sector agencies who are at the early stages of AI adoption.

However, he cautioned that reliance on external experts should be carefully managed to ensure that knowledge transfer occurs internally, and that long-term capability is built within public sector finance teams.

“By fostering a collaborative environment where finance professionals, data scientists, and AI specialists work together, governments can create a resilient and adaptive approach to AI implementation,” he explained.

Aside from industry experts, governments should also engage with the ecosystem broadly, including other regulatory bodies and academia, to support the development and use of AI that aligns with both public interest and compliance needs, he noted.

Whole-of-government approach needed

Brisbourne pointed to a broader range of AI use cases in public finance, and explained why a whole-of-government approach is needed to tackle the challenges that come with the widespread deployment of AI in the public sector.

“Beyond just catching mistakes and fraud, AI can help make budget planning smarter.

“We’ve seen local authorities, for example, using machine learning (ML) to predict which properties are most likely to have serious building code violations, helping them allocate their limited inspection resources more effectively.

“Meanwhile, national governments have explored AI-enabled data sharing across government departments, cutting down on duplicate reporting and making it easier to track public spending in real-time,” he explained the emerging use cases.

But while AI is incredibly powerful, it is not infallible. That fact highlights the need for robust human oversight and clear governance frameworks to ensure AI tools are being used responsibly in public finance.

As well as why a coordinated effort across government agencies and regulatory bodies is needed to manage AI-related risks, which can range from cybersecurity threats, misleading or false information, data privacy concerns and ethical dilemmas, he said.

To achieve this coordinated effort, ensuring workforce readiness to tap AI effectively and responsibly, as well as setting boundaries for AI use through an adaptive and agile governance framework would be key moving forward.

“Frontline workers would need to be able to assess these risks which are the product not just of the AI system but also of how we use the system or information provided,” he explained.

For public sector leaders looking to prepare for this future, investing in AI literacy and talent development should be a top priority.

Upskilling efforts for public finance professionals should include practical applications of AI tools, ethical AI and risk governance, and continuous professional development to ensure that finance professionals remain equipped to navigate the evolving AI landscape.

Above all, governments must foster public trust by demonstrating a commitment to responsible AI governance, ensuring that AI-driven financial systems serve the best interests of society.

ACCA offers a suite of free webinars and paid courses that are specific to AI in public finance.