Most of us would be thrilled with a 97% rating for an AI-driven SoA review, but we’d be less thrilled if a human subsequently reviewed the file and found the advice inappropriate.

We’d be even less thrilled to read the file and realise they’re right. Perhaps the AI didn’t review the full file, but the Fact Find and file-notes clearly flagged material issues. The AI, in this case, didn’t just miss the problem; it offered false confidence. And suddenly that 97% score isn’t comforting. It’s terrifying because it forces you to face the fundamental question: what else are you letting through your monitoring and supervision net just because the system provides mathematical assurance?

This isn’t an indictment of either the service provider specifically or AI generally; it’s rather constructive criticism of those Licensees that treat tools as a substitute for expertise, due diligence and judgement.

The hype around AI is deafening, but the real question for financial services isn’t how to use it. It’s whether your people, culture and systems are ready to make it work.

The 2025 EY Work Reimagined Survey delivers a sobering insight for licensees, advisers and Responsible Managers: although 88% of employees now use AI tools at work, only 28% of organisations are positioned to realise meaningful business value from that usage. Technical barriers don’t cause that gap. It’s caused by culture, leadership, and underinvestment in people.

For those operating in a highly regulated sector like financial advice, the message is clear: adopting AI without addressing workforce strategy is more likely to introduce operational and compliance risks than to deliver productivity benefits.

What is the AI Illusion?

It’s tempting to believe that artificial intelligence will deliver quick wins and immediate and lasting advantages, from writing tools and risk management to workflow automation. And at the task level, it often does. Advisers may use generative AI for document drafting or research. Compliance teams may automate monitoring or analysis. AI is a “good tool but a poor master“; when these technologies are layered onto underperforming teams or rigid processes, or used to replace rather than augment expertise, sub-optimal results are the best outcomes you can hope for.

EY’s data confirms that even where AI use is high, few organisations achieve real uplift in value or productivity. General observations may not account for your own exceptionalism, but the reality is that technology can’t fix cultural fragility, disengagement or poor systems. In fact, it often makes those weaknesses more visible and even more problematic.

Sure, well-run firms can stumble with AI, and other firms can run into problems without using or relying on AI, but not to the same degree (97%) of delusional false-confidence.

While we’re being honest, we should acknowledge that AI doesn’t fail in isolation; it fails in context. That’s why the real challenge isn’t technological, it’s organisational. The illusion of AI competence often masks fragility in people systems: thin training, misaligned incentives, or leadership that treats capability as an HR issue rather than a compliance risk. This brings us to what EY calls the “Talent Advantage”.

Why is Talent Strategy a Compliance Issue

Exposing the placebo effect is one thing, but EY’s survey also introduces the “Talent Advantage”. This is a useful concept for describing a state in which organisations align people and systems. Firms with a Talent Advantage are more likely to see business value from AI, and are less exposed to regulatory, conduct or operational risk.

For compliance professionals, this reframes AI not just as a technical implementation but as a governance and risk issue. A position reiterated in ASIC’s REP 798 “Beware the gap: Governance arrangements in the face of AI innovation”.

 If you’re automating systems without upskilling, implementing controls without critical oversight, or cutting headcount while overloading the remaining staff, you’re not future-proofing your business; you’re embedding systemic failure.

Compliance leaders should ask:

• Have our AI tools changed adviser conduct, and have we updated procedures accordingly?
• Are we investing in upskilling, continuous learning, and actual capability building?
• Are our reward structures (including variable remuneration) aligned to desired conduct and outcomes?

Does Quit-Intent Increase Regulatory Risk?

One of the more interesting observations in the survey is the rise in “quit-intent”, the number of employees planning to leave within the next 12 months. This is particularly high in firms without a Talent Advantage. The risks go beyond commercial disruption because high turnover in advice firms (or even just moderate turnover in critical roles) can lead to supervision gaps, quality failures, and breaches of licensee obligations under s912A(1)(a)-(ca) of the Corporations Act 2001 (Cth).

This isn’t something to note and ignore. If these observations apply to financial services, the increasing number of staff with “go-bags” may represent an emerging compliance risk for Responsible Managers and Licensees, who are expected to maintain “adequate resources” and “competent staff” to provide services efficiently, honestly, and fairly. AI can paper over workforce deficiencies in the short term, but concealing rather than addressing them exposes the business to increased regulatory risk.

Lessons for Licensees and Advisers

If you’re running a professional advice business, here are four practical lessons drawn from the EY research:

1. Don’t Over-Index on Tools. Build Capability.

AI can extend human performance, but it doesn’t replace foundational skills or judgment. Compliance teams should validate that AI tools enhance adviser conduct, such as best interest duty (s 961B) or recordkeeping. AI might not hallucinate, but it confabulates, and this can be profoundly damaging to a business that relies on and is vulnerable to AI controls.

2. Revisit Training and Supervision Policies

Upskilling can’t be optional. Firms should move beyond checkbox CPD and invest in applied, ongoing capability frameworks that reflect how AI changes roles and workflows. This may involve integrating learning into performance reviews, ensuring that supervisors understand new risk indicators, and confirming that management understands the limits and weaknesses of system solutions.

3. Monitor Training and Supervision Policies

Culture remains one of the most significant indicators of risk. If staff feel unsupported, undertrained or micromanaged, their likelihood of disengagement and error increases. Regular pulse checks, exit interviews and sentiment analysis should inform compliance reviews.

4. Reward What You Want to See

Many firms still incentivise activity rather than outcomes. The EY report shows that misaligned reward systems undermine adoption and trust. For advisers and support staff alike, compensation should reflect quality, not just quantity — including adherence to compliance obligations and conduct expectations.

How Should Compliance Leaders Handle AI?

The AI wave is already reshaping advice and compliance, but whether it sharpens your edge or exposes your blind spots depends on what it breaks. Technology doesn’t fix culture or capability; it amplifies what’s already there. That 97% review score felt reassuring, until it wasn’t. The truth is that when AI is layered on top of shallow governance, fatigued teams, or rigid processes, it doesn’t protect you; it simply papers over risk and delivers false confidence at scale. The hard part isn’t adoption. It’s readiness. And in financial services, readiness means more than having the right tools; it means having the judgment, systems and culture to make those tools work without undermining trust.

AI is pregnant with possibilities, and FOMO is real, but the benefit of AI is neither inevitable nor guaranteed. Try to resist the AI evangelists that infect LinkedIn and PD Days with the unfounded promises of guaranteed compliance and limitless profits.  The firms that succeed will be those that treat people strategy as central to risk and performance, not an HR afterthought. This is particularly critical in financial services, where trust, judgment and ethical conduct are not optional but fundamental to the value proposition.

Remember that your adoption of any AI tools is accompanied by documented policy updates, revised controls, and (most importantly) cultural alignment. AI is not a plug-and-play solution; it’s a lever. It magnifies what your firm already is, at speed and scale.

So, if you want AI to enhance compliance and advice quality rather than accelerating your risk, then your governance, people, and processes must be ready to support it.

If you’re serious about using AI without sleepwalking into a breach, start with your governance, not your gadgets. Assured Support can help you translate ASIC’s expectations and EY’s “Talent Advantage” into practical AFSL compliance arrangements like independent file reviews, AI-aware monitoring programs, culture diagnostics and tailored training for Responsible Managers and advisers. Talk to us before you automate, so your systems, people and controls are ready for the tools you’re buying.

If you enjoyed this article, you might also like:

• Adoration of the mAgI (Why do financial planners need to be careful with AI?)
• AI in Financial Advice: Efficiency Gains, Compliance Risks, and Cognitive Costs
• The Hidden Risks of AI: ASIC’s Review of Licensees’ Embrace of Artificial Intelligence

Frequently Asked Questions