AI in Finance: Hype, Reality, and What’s Next
AI is no longer just a buzzword in finance. It is already reshaping operations, risk management, and reporting — but the full promise hasn’t yet been realised everywhere. Here are five realities finance leaders must understand today.
1. Automation Delivers Value — But Only in the Right Places
The global AI in financial services market was valued at USD 38.36 billion in 2024 and is projected to reach USD 190.33 billion by 2030, growing at a 30.6% CAGR (MarketsandMarkets, 2024).
According to McKinsey’s State of AI 2025 survey, 78% of organisations now use AI in at least one business function, up from 72% in 2024.
Automation delivers real efficiency in repetitive, rule-based, data-rich tasks such as invoice processing, reconciliations, and regulatory reporting — all ideal for machine learning.
However, areas requiring context, judgement, or nuance (e.g., forecasting, strategy, or investment decisions) still depend heavily on human oversight.
2. Data Quality Determines Outcomes
The effectiveness of any AI tool depends on the quality and governance of data.
A 2025 Pigment Research survey found that over 60% of finance leaders cite fragmented or inconsistent data as the biggest barrier to AI adoption (Pigment, 2025).
Organisations with clean, well-structured data pipelines are already seeing shorter close cycles and more accurate forecasts, while others face unreliable results and sunk investment.
In short: data maturity determines AI ROI.
3. Generative AI Is Helpful — But Not a Substitute
Generative AI (GenAI) tools can draft reports, summarise data, and generate first-cut insights in seconds — tasks that once consumed analysts’ days.
But accuracy remains an issue: a Deloitte 2025 report notes that 70% of GenAI outputs in finance workflows still require human review.
The takeaway? Treat GenAI as a junior assistant, not a replacement. It speeds up work but doesn’t yet remove the need for domain expertise, contextual thinking, and accountability.
Sources: (Deloitte, 2025), (Pigment, 2025)
4. Governance and People Make the Difference
As AI moves from experiment to core, governance, explainability, and regulatory alignment become critical.
The U.S. Financial Stability Oversight Council (FSOC) in its 2025 report warned that increasing reliance on AI in trading, risk scoring, and credit underwriting presents “new concentrations of systemic risk.”
Leading financial institutions are responding by investing in:
- AI oversight frameworks that clarify accountability for outputs
- Bias detection and model-risk management programs
- Upskilling in AI literacy across finance, audit, and compliance teams
Source: (FSOC Annual Report, 2025), (RGP, 2025)
5. A Big Push: Automating the Analyst Grind
In early 2025, the Financial Times revealed that OpenAI has been working on “Project Mercury”, a program designed to train models on banking-analyst tasks such as financial modelling, IPO prep, and restructuring analysis (FT, 2025).
The company reportedly hired over 100 former investment bankers from firms including J.P. Morgan, Goldman Sachs, and Morgan Stanley, paying around USD 150/hour to simulate real analyst workflows and feed this data back into its AI systems.
The goal is to automate the “grunt work” — building Excel models, adjusting pitch books, and preparing sensitivity analyses — that junior analysts typically spend 70–80 hours per week completing.
While this doesn’t eliminate human analysts, it signals a future where junior talent focuses on insight and client engagement rather than mechanical number-crunching.
Conclusion
AI in finance is real and valuable — but its impact depends on execution.
The firms that win will be those that combine:
- strong data foundations,
- thoughtful governance, and
- empowered, AI-literate people.
And increasingly, those that strategically automate routine work will create space for human analysts to focus on higher-value thinking.
For them, AI is not hype — it is a tool for sharper decisions and smarter growth.
