AI and Compliance Careers, Part II

You might remember that the other week we had a post tracing the traditional career path of compliance professionals, and how that path might change as artificial intelligence seeps into corporations and new risks keep crowding onto your plate. Today we have some reader feedback that raises some great points about the future of compliance careers.

The feedback comes from reader F.D., an ethics officer at a large U.S. defense contractor. We’ll start with his views on how AI might affect individual compliance officers’ workloads:

Your prediction about AI piling more work on to the manager’s plate is a highly likely outcome. Case in point is how process automation reduced the need for admin support jobs considerably over the years. In the past, once one made it to “manager” level, one would not answer calls, make travel arrangements or file expense reports; you had support staff to do that.

Today, thanks to process automation (voicemail, travel booking portals, automated expense reporting), managers do all of that for themselves and monitor their team’s compliance with those processes too. Directors might have access to shared admin support, and only VPs or higher get dedicated admin support.

If we expect a similar transformation with AI, it’s not outside the realm of possibility that in the compliance field, this might reduce the need for analyst roles and pile more work on the managers (initially). We will need analysis as usual, but it will be delivered by AI in highly customized forms and formats, and may even come with summary interpretation of the data.

F.D. raises good points. They’re similar to what I’ve speculated before, that perhaps AI won’t free low-level employees from scutwork so that they can focus on more high-skill challenges. Perhaps it will burden high-level employees with more oversight duties as they double-check the results that AI is providing.

That is, instead of AI unshackling the bottom of the org chart so those folks can reach upward, perhaps it will reshackle the middle rungs of the org chart, where you keep reaching backward to refine AI’s almost-but-not-quite output. (That was one of the findings from the Workday survey I cited in another post last week: many employees aren’t reaping efficiency gains from AI because they’re so busy confirming and fine-tuning its output.)

What Compliance Officers Will Do

Then F.D. moved to another point about what compliance officers will actually do when corporations are infused with AI agents:

In my mind, a question more critical than where compliance reports (into legal or anywhere else) is what the compliance function will work on. In an organization heading toward an AI-centered enterprise, AI agents will be able to provide interpretive guidance on company policies to employees or to identify and investigate non-compliance. Not to mention AI’s already existing capability to create interactive training modules from policy documents or to develop communication pieces on key compliance topics.

One critical skill that compliance professionals possess and practice everyday is their ability of working across the functions to dig into issues and map where the processes failed and how people took advantage of those weaknesses. Although corrective actions are taken, those actions are often band-aids fixing the damage from that single incident, rather than deeper, more systemic corrective actions.

If compliance professionals can be empowered and assigned to implement systemic corrective actions working across the enterprise, it may be a while before AI can come up to par there.

Again, good stuff from F.D. AI will automate lots of policy development, training, communications, and regulatory change management work that human compliance employees do now. AI will also be a huge help for transaction monitoring and anti-fraud programs, flagging suspicious transactions that could be money-laundering, embezzlement, corruption, and so forth — stuff that lower-level compliance analysts do now.

Does that mean AI will eventually lead to mass unemployment and extinction for compliance officers? I still doubt it.

First, let’s appreciate the difference between AI helping companies to “do compliance” more efficiently, and AI helping companies to do more compliance.

For example, if AI can help financial firms do transaction monitoring more efficiently and more accurately, that’s great; because right now, most firms’ transaction monitoring systems are terrible. They need that AI-enhanced help, to stop wasting human time on false positives while the actual bad stuff sails by undetected. This will be even more true in the future, as AI-enhanced frauds and scams become more prevalent.

Along similar lines, AI will be a godsend for smaller companies because right now they have no substantive compliance capability at all, but they are hitting compliance obligations earlier in their natural lifecycle. So they’ll need that AI help to navigate regulations, contractual obligations, and audit requirements that they never would have encountered in, say, 1995 or 1970.

And Don’t Forget Complexity

We also have F.D.’s point that humans can “work across the functions” and recommend deeper, more systemic corrective actions.

That’s important. AI is good at executing certain pre-established tasks, but it’s not good at designing whole new processes — especially when those processes involve humans, who have ambitions and fears and nobility and wisdom that AI doesn’t, and (hopefully) never will. AI can’t look across multiple parts of an enterprise and rewrite an org chart or concoct new delegations of authority.

Of course, an implicit question here is whether senior management teams will actually care enough about ethical conduct and prudent risk management to let compliance officers identify and implement those deeper, systemic corrective actions. Maybe they won’t. Maybe some management teams will be content to pile one band-aid atop another, addressing the immediate incident rather than the root cause.

But when management teams do care enough to say, “Hey, can we step back and look at the bigger picture, to see whether there’s a better way to do this in the future?” — that’s where compliance officers will be able to excel, and AI could even help you to achieve that excellence.

Anyway, that was F.D.’s perspective. If you have thoughts you want to share about the future of the compliance profession, you can always drop me a note at [email protected]! A human response is guaranteed.