Three critical lessons from Workiva's Amplify event on AI governance in enterprise workflows

The shift in enterprise AI governance is now focused on the underlying data, processes and guardrails that shape what automated systems can accomplish. When AI agents handle consequential decisions across finance, audit, compliance and sustainability functions, organizations must ensure outputs are anchored in business reality, linked to verifiable sources and subject to transparent accountability structures.

TheCUBE's coverage of Workiva's Amplify event examined the governance requirements emerging as AI transitions from suggesting actions to performing them. Connecting information with surrounding business context is essential for trustworthy automation, according to theCUBE Research's Krista Case.

What I'm looking at as Workiva's bigger opportunity is creating that context and being that connective tissue across those disparate platforms and bringing in that more tribal knowledge … to make sure that as we are starting to ask AI to actually execute functions, it can be trusted because it does increasingly have that context behind it. They are historically very strong in things like compliance and reporting and very regulated and audited workflows. They are beginning to make some steps into areas like sustainability.

Krista Case, theCUBE Research

Case and co-host Alison Kosik interviewed executives and practitioners at the Amplify event during an exclusive theCUBE broadcast on SiliconANGLE Media's livestreaming platform. The discussion examined how AI agents are reshaping accountability structures, how Workiva is reimagining governed workflows and how organizations are evaluating which business processes are suitable for automation.

https://www.youtube.com/embed/XS24OU9KdGo?feature=oembed

Insight #1: Enterprise AI governance must define the limits of agent autonomy

As AI agents evolve from supporting workers to executing regulated tasks, governance requirements expand beyond data quality alone. Organizations require an auditable trail that enables accountability once a task concludes, Case explained.

We need to understand things like: Where did this information and … these insights come from? Who approved a particular action? Can you maybe trace what happened if an AI agent is taking an action on your behalf? 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.

Krista Case

A central design challenge involves determining how much autonomy to grant an agent while preserving meaningful human control. The appropriate balance depends on risk profile and specific business context, with monitoring and escalation mechanisms helping organizations retain oversight.

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. We need to kind of make decisions around when AI is assisting a human, when it can execute on its own [and] when it needs approval. I do think that some of these boundaries are still being defined, and I think they're going to evolve and change over time, especially as the business use cases evolve.

Krista Case

https://www.youtube.com/embed/8Kafp1RFE4Q?feature=oembed

Insight #2: Governance and workflow redesign must advance together

Workiva's 2026 Midyear Executive Benchmark Survey reveals a significant tension in enterprise AI governance: a disconnect between executive confidence and organizational readiness. Eighty-four percent of executives surveyed expressed at least moderate confidence in AI-generated outputs for annual reports without human review, yet only 11% reported their data quality met the standards required for AI deployment. Financial controls grounded in four foundational elements—data origin, personnel involvement, data transformation and review-approval processes—can strengthen accountability in AI-assisted reporting. This accountability becomes even more critical as AI accelerates workflows, according to Steve Soter, vice president and industry principal at Workiva.

It actually makes it even more important because accelerating a process, if you don't have it grounded by … those four things, then speed doesn't become an asset. It really becomes a liability. It becomes a risk.

Steve Soter, Workiva

When this foundation exists, governance can enable broader workflow transformation rather than impede it. Agent Studio offers business users a no-code interface for constructing and tailoring agents to their specific workflows, with early adopters building functional agents in just days, according to Kim Huffman, chief information officer of Workiva. Workiva is applying an outcome-focused methodology internally, where employees across departments establish a shared objective and then identify where agents can contribute to the work.

I think AI is a fundamental technology that's going to reshape the landscape of technology portfolios in every organization. When you think of it as a bolt-on capability, you're probably going to get bolt-on outcomes.

Kim Huffman, Workiva

https://www.youtube.com/embed/mYlG02ySzzk?feature=oembed

Insight #3: AI readiness must be evaluated process by process

At HP Inc., sustainability information serves regulatory obligations, fulfills customer inquiries and informs product strategy across operations spanning more than 180 countries. This geographic scope makes data consistency, verification and formal approvals critical, according to Jen Huffstetler, senior vice president, chief sustainability officer and global market access at HP, and Kristen Siemen, senior advisor at McKinsey & Company Inc.

If I think back to my time at General Motors, we ran the exact same processes for sustainability data and reporting that we did for financial. So, internally, our reviews, our assurance, the process and the sign-offs were all consistent.

Kristen Siemen, McKinsey & Company Inc.

Uniform standards do not mean all processes have equal readiness for automation. At Accel Entertainment Inc., initial deployment focuses on Securities and Exchange Commission filings and enterprise resource planning systems where data lineage is well understood and auditors can participate early. Evaluating individual disclosures by risk level before determining AI suitability is essential, according to Christie Kozlik, chief accounting officer at Accel.

What is the input that I am feeding into this AI? What is the output that I'm expecting? What is the root cause that I'm trying to solve for? Knowing those inputs, outputs and expectations becomes so important.

Christie Kozlik, Accel Entertainment Inc.

Readiness assessment also shapes deployment velocity. At Newell Brands Inc., internal audit distinguishes lower-risk implementations, such as agents handling customer order status inquiries, from higher-risk agents managing fixed-asset accounting. Audit teams collaborate with business and technology groups as deployments scale, according to Keri Tracy, vice president and chief audit executive of Newell Brands.

https://www.youtube.com/embed/U0OKu4tpLYU?feature=oembed

https://www.youtube.com/embed/videoseries?list=PLQ4i1rVrbKBE

I think a lot of companies don't understand the concept of change management well enough to understand that if you don't bring the people along with you, you could have the best plan, you could have the best technology, and it won't work. I think a lot of transformations fail because they don't bring the people along.

Keri Tracy, Newell Brands