As Gen Z flocks back to accounting, EY is investing $100 million in bonuses for employees who prove they have human skills

By Gammatek ISPL, Industrial Compliance Analyst at Gammatek ISPL

Published: September 2026 | 10 min read

Author block: Gammatek ISPL advises manufacturing, chemical, and pharmaceutical plants on compliance strategy and safety systems at Gammatek ISPL, drawing on direct work implementing compliance software across + industrial facilities. Sources for this article include public reporting from Fortune, CBS News, CFO Dive, and Accounting Today, alongside Gammatek's own client observations. This is independent analysis, not sponsored by EY or any firm named below.

As AI takes over routine compliance checks, the human judgment behind an audit is becoming the part that actually matters most.

Why This Should Matter to Anyone in Industrial Compliance

This week, Ernst & Young announced it's <cite index="7-1">investing $100 million to reward U.S. employees who develop "future-focused skills, advance the firm's culture, drive innovation, and deliver exceptional client service"</cite>. Individual awards can reach $25,000. The headline reason is simple: as AI takes over the repetitive parts of accounting, the firm needs to prove — to clients and to itself — that the human judgment layered on top of that work still justifies the price tag.

If you run compliance, safety, or quality assurance for a manufacturing, chemical, or pharmaceutical plant, this isn't a story about accountants. It's a preview of a decision your own organization is going to face — probably sooner than you think. AI is already automating large parts of industrial compliance work: log review, anomaly detection in sensor data, checklist verification, even first-pass audit report drafting. The open question EY is answering for its own industry is the same one compliance leaders in manufacturing need to start answering now: when AI does the routine checking, what exactly are you paying your compliance team to do — and how do you prove that value to regulators, auditors, and your own leadership?

What EY Actually Did, and Why It's Not Just a PR Move

The EY program isn't a blanket raise — it's structured recognition. <cite index="1-1">Individuals can earn spot awards up to $500, while individuals and teams whose work makes a material difference to the firm can receive cash awards up to $25,000</cite>. The skills being rewarded are specific: <cite index="1-1">business acumen, judgment and adaptability, and experimenting with technology to drive innovation and improve client services</cite>.

What's notable is why EY felt compelled to formalize this now. <cite index="1-1">AI has made accounting more attractive to Gen Z entering the profession as it strips out the "boring" manual tasks, but that same shift creates a harder problem: convincing clients that the human expertise behind the work is still worth paying for</cite>. In other words, AI didn't just change the workload — it changed the sales pitch. Clients can increasingly ask, reasonably, "if AI does most of this, why am I still paying senior rates?" EY's answer is to make the human contribution visible, measurable, and rewarded, rather than assumed.

That's the exact same conversation compliance leaders in manufacturing are going to have with plant managers, regulators, and CFOs — just a few years behind.

The Industrial Compliance Version of This Problem

Here's where Gammatek's own client work adds something the accounting story doesn't cover: what this looks like on an actual plant floor.

Over the past two years, we've watched AI-assisted monitoring move from a novelty into a baseline expectation at the facilities we work with. Sensor-based anomaly detection, automated log review, and first-pass checklist verification — the equivalent of accounting's "routine" work — are now standard in a growing share of new compliance software deployments, FixitX included. That's a genuine efficiency gain. Plants that used to spend hours of staff time on manual log review now get flagged anomalies in near real time.

But here's the pattern we consistently see, and it maps almost exactly onto what EY is responding to: automating detection doesn't reduce the need for judgment — it relocates it. A flagged anomaly on a chemical plant's temperature sensor still needs a qualified person to decide whether it's a sensor fault, a process drift, or a genuine safety risk. A compliance software platform can surface the pattern; it can't take regulatory accountability for the decision. That accountability — the signature on the audit report, the judgment call under ambiguity, the decision to halt a line versus flag-and-monitor — is the part that doesn't automate away, and it's exactly the layer EY is now paying a premium for in accounting.

Implementation consideration: for plants currently evaluating or expanding AI-assisted compliance tools, the practical question isn't "should we automate detection" — most already are, or should be. It's whether your compliance staffing and training model has caught up to the fact that their job has shifted from finding problems to judging problems. Teams still resourced and trained primarily for manual log review are increasingly mismatched to what the role actually requires now.

A Simple Framework: Where Judgment Still Has to Sit

Based on what we've seen across client deployments, compliance tasks generally sort into three buckets:

Task type	Example	Can AI own it?
Detection	Sensor anomaly flagging, log pattern review	Yes — this is where automation adds the most value with the least risk
Interpretation	Deciding whether a flagged anomaly is a real risk or noise	Partial — AI can support with historical pattern data, but shouldn't decide alone
Accountability	Signing off on an audit, deciding to halt production, regulatory reporting	No — this needs to stay human, both practically and often legally

The mistake we've seen a handful of plants make is treating the whole compliance function as one bucket — either resisting automation entirely (falling behind on efficiency, same criticism EY implicitly leveled at parts of traditional accounting) or over-trusting automation into the accountability layer, which creates real regulatory and safety exposure.

What This Means Practically, Right Now

A few concrete takeaways for compliance and plant safety leaders reading the EY story and wondering what it means for their own operation:

1. Audit what your compliance team actually spends time on today. If most of their hours are still going to manual detection work AI tools now handle well, that's a resourcing mismatch, not a headcount problem — the fix is retraining toward interpretation and judgment, not necessarily cutting staff.

2. Make the judgment layer visible, the way EY just did. EY's move was as much about proving human value to clients as to employees. Compliance teams should be documenting and communicating the judgment calls they make — not just the audits they pass — so leadership and regulators see the value that isn't captured by "the software flagged zero issues."

3. Don't let automation quietly creep into the accountability layer. It's tempting to let a well-performing detection system's outputs go unreviewed over time. That's exactly the gap that creates regulatory and safety risk — the software found the pattern, but nobody with real accountability actually looked at it.

4. Expect the same generational shift EY is describing. Just as <cite index="1-1">AI has made accounting more attractive to Gen Z by removing some of the "boring" tasks</cite>, the same is likely for industrial compliance roles — younger hires may be drawn to compliance work specifically because AI has stripped out the tedious log-review grind, leaving more of the analytical, judgment-driven work that's actually interesting. That's worth factoring into how plants recruit and train compliance staff going forward.

Where Gammatek Fits Into This Shift

This is exactly the design principle behind how we've built our compliance platform: automate detection aggressively, but keep accountability, sign-off, and judgment calls explicitly in human hands, with a clear audit trail showing who reviewed what and when. It's not AI replacing your compliance team's judgment — it's AI clearing the routine work off their desk so the judgment they're actually paid for gets more of their attention, not less.

If you're evaluating how AI-assisted monitoring fits into your plant's compliance program — or want a clearer picture of where your team's time is currently going — [see how Gammatek's compliance platform is built to keep human judgment at the center →