Workiva Inc.'s Amplify event surfaced a clear signal this week: AI governance is no longer a back-office compliance checkbox. It is moving directly into the workflow, driven by the reality that AI agents are now performing regulated work - not just suggesting it. For HR and management leaders, the message was blunt. If a finance AI agent can execute a transaction, the organization must be able to trace where the information came from, who approved the action, and what the agent did on the company's behalf.
The shift puts new pressure on data foundations that have been fractured for years. Krista Case, principal analyst at theCUBE Research, said the problems are not new - but the stakes are. "I think it's commonly understood that our AI is only as good as the data that it's built on," Case said during an interview with host Alison Kosik at the event. "What we talked about more specifically here at Workiva Amplify was the fact that if we have fragmented data stores, if we have inconsistent definitions and inconsistent ownership over data, these are not necessarily new problems that were created as a result of the enterprise adopting AI. They're problems that existed before."
When agents act, accountability gets harder
Fragmented data becomes dangerous when AI agents can turn flawed information into decisions at speed. In reporting, audit, and compliance - Workiva's core domains - plausible output is not enough. Case pointed to the need for substantiation. "We need to understand things like where did this information and where did these insights come from, who approved a particular action, and can you maybe trace what happened if an AI agent is taking an action on your behalf," she said. "When you think about finance and these other regulated processes, it's not really good enough for the action or the response to just look plausible. We have to really make sure that it can be substantiated."
That requirement forces a hard conversation about human review. If a person must approve every AI action, much of the speed benefit disappears. Case described the tension directly. "What's interesting is that going back to the conversation around speed, if a human has to review and approve every action, then really the whole point or much of the value is moot," she said. The emerging answer is risk-based controls - deciding when an agent assists, when it executes independently, and when it must stop for approval. Those boundaries are still being drawn and will shift as business use cases mature.
Governance becomes a workflow design problem
For HR and management teams, this is not a distant IT concern. The same governance questions will apply to AI agents used in compensation analysis, performance review drafting, onboarding workflows, and workforce planning. If an agent flags an employee for a retention risk score or recommends a salary adjustment, someone needs to know what data drove that conclusion and who - or what - made the call. Monitoring and exception handling become central skills, not afterthoughts.
Workiva's event highlighted that governance tools must live inside the platforms where work happens, not in separate compliance dashboards. That principle applies across functions. As organizations build or buy AI Agents & Automation capabilities, they need the ability to reconstruct an agent's actions, inspect data lineage, and adjust approval thresholds without slowing work to a crawl. The design challenge is making governance fast enough to keep up with the automation it oversees.
Why this matters for HR and management
AI governance is becoming a people-management skill. HR leaders who understand data lineage, approval workflows, and agent traceability will be better positioned to shape policies before an incident occurs. The alternative - bolting on controls after an agent makes an embarrassing or costly error - is far more expensive. For managers evaluating AI tools in hiring, performance, or compliance, the key question is no longer "Does it work?" but "Can we prove how it worked?" Building that capability now, starting with clean data ownership and clear approval rules, is the practical takeaway from Amplify this year.
Your membership also unlocks:
