I was recently invited to serve on a Panel at UNISA’s Symposium on "Continuous Auditing with Artificial Intelligence". Naturally, my focus was AI in the context of accounting and reporting. The photo is courtesy of George Higgins – a fellow panel member.

AI has the potential to be a game changer in many aspects of accounting and reporting. Sadly, the implementation is slow. Apart from the perceived complexity of AI, finance professionals may think they will be replaced by AI. In my view, AI is useful for routine transactions. However, because accounting is a social science and requires the application of judgement, human accountants, auditors and others continue to play a critical role. Here are a few thoughts on where AI could be useful in the public sector.

Blockchain
Blockchain is a decentralised, distributed, digital ledger that records transactions across a peer-to-peer network.  Each transaction is grouped into blocks, which are linked together chronologically, forming an unbreakable chain. Transactions cannot be deleted or changed, and as a result, Blockchain creates a shared, tamper-proof record of transactions. Changes can only be made if the participants in the network agree. This would be ideal for accounting records that have a level of permanency or “master data”, for example, land and property records, beneficiaries of grants, etc. A concern in using Blockchain for these transactions could be security. As Blockchain uses mostly public networks, whatever is recorded will be available to all. While there are private networks, this compromises the benefit of having a decentralised network where a group of parties is required to change a record.

Smart contracts
Smart contracts use Blockchain technology and are self-executing contracts that execute automatically when certain conditions are met. Contract management is notoriously bad in government leading to non-payment of service providers, or paying for goods and services not actually rendered. While smart contracts are not legally recognised in the South African legal framework, this could be an area of value for the sector.

XBRL - (Extensible Business Reporting Language)
XBRL standardises financial reporting using a specific reporting framework and make reports available in a computer-readable format. This aids standardisation and making information available more widely. XBRL requires transactions and the associated data to be tagged using the relevant reporting framework (at present, only IFRS). AI could be used to extract data from records, reports etc. and help the tagging process. AI could assist with ensuring the credibility and completeness of the data used in XBRL. While XBRL is not available for Standards of GRAP or IPSAS at present, other automation tools could benefit from using AI.

While I am sure there are several more benefits – and risks – one cannot ignore technology and how it could be used to make our work more efficient and add value to citizens. Happy to hear your thoughts on other AI benefits and risks by posting these on our LinkedIn or Facebook pages.