Australia Banks: AI in Banking: Will agentic AI be an agent for guidance?

研报英文原文证据摘录

Australia Banks: AI in Banking: Will agentic AI be an agent for guidance?

Australia Banks

21 July 2026 Citi Research

Source: Citi Research, Evident Insights

Industry research indicates that broader adoption is increasing, but the skew

amongst adopters is seeing increasing bifurcation. Unsurprisingly, higher

performance in AI indices correlates with higher AI deployment and use cases

across various banks. As is the case in Australia, it is important to note that to date,

broader deployment has not correlated with financial outcomes as the majority of

cases are largely test cases. However, adoption is broadening out at pace. As we

approach our fifth year of GenAI, global industry research suggests that the

strongest adopters of AI are now starting to pull ahead from peers, aided by

investment in talent and innovation.

AI, as a broad church of related technology, has seen mass adoption at a pace not

seen in other technology. AI has been in place for some time, and consequently

aspects of it are well entrenched across the financial services landscape. Classical

machine learning (or narrow AI) has been broadly adopted, be it in fintechs or

traditional institutions. Generative AI (GenAI) has been in place since 2022, and

has seen similarly rapid take-up and broad adoption. Agentic AI remains the next

frontier, with penetration relatively lower amongst traditional institutions.

However, conversations with the industry and the banks suggest this is where they

see the most economic value capable of being delivered.

Figure 2. Active AI adoption by AI type and stakeholder group

AA

“Active adoption” includes piloting, scaling or transforming

Source: Cambridge Centre for Alternative Finance

… but more tangible business impacts could arise as agentic AI penetration lifts.