Outcome-centric governance is the key to enterprise adoption of GenAI

The market sentiment on adopting generative artificial intelligence (GenAI) appears to have shifted. Organizations are no longer just trying to understand the technology better and dabble with some proofs of concept (PoCs)—they are looking to scale out deployments. Yet, the fundamental challenge for businesses remains that if they are serious about deploying AI, they must prepare their organizations for a different way of working. A further challenge is that they don’t always know how GenAI affects the outcomes of their activities and processes. Therefore, the governance of these issues and challenges requires innovative approaches. EY’s partnership with ServiceNow to provide solutions for GenAI compliance, governance, and risk management is a compelling proposition to advance all those discussions.

Beware of the hype, and stay focused on the fundamentals

Let’s start with the small print. The partnership includes the EY AI Governance and Compliance solutions, which provide businesses with capabilities such as AI discovery and inventory management, policy management and implementation, risk tiering, and automated monitoring. Conversely, EY will use ServiceNow GenAI capabilities internally across IT and HR business functions, including IT Service Management (ITSM) PRO+ deployments, HR Service Delivery (HRSD) PRO+, Now Assist, and GenAI tools.

The fundamental value proposition of EY AI governance solutions is to treat GenAI as an asset that a configuration management database (CMDB) mindset can help manage. The CMDB is the cornerstone of the ServiceNow platform. It offers a comprehensive and dynamic inventory of all an organization’s IT assets and services, mapping their relationships and dependencies. Simply put, the CMDB provides a single pane of truth, enhancing visibility, enabling informed decision-making, and fostering operational efficiency across an organization. What stands out here is EY’s holistic approach, which is neither just looking at technology issues nor getting lost in slippery statements on AI ethics.

EY’s ServiceNow-specific approach is amplified by its broader methodology on AI governance. It focuses on patterns, not use cases, to achieve a faster time to value. These could be functional patterns, such as summarization, insights, translations, predictions, and Microsoft Copilots. It could also retrieve augmented generation (RAG), knowledge graphs and ontologies, multi-model systems, and AI agents. As you would expect from EY, the go-to-market is consultative, with discussions on business cases being central in the early phases of engagements. This approach is a refreshing demarcation from the capabilities-centered hype around GenAI.

The ServiceNow ecosystem is on overdrive on GenAI

Regarding innovation, ServiceNow’s executives have a knack for doing two things exceptionally well that, at face value, appear to conflict. On the one hand, Bill McDermott is a master of impressing financial analysts with eloquent descriptions of new target addressable markets, such as automation and AI, that appear to be pushing the innovation envelope. On the other hand, releasing new features and capabilities is done carefully, if not conservatively. As seen with its most recent release, Washington, most AI innovations are around Copilot use cases, but not scenarios other ISVs wouldn’t have done as well. ServiceNow leverages GenAI to drive efficiency across its platform, but it is leaving more innovative use cases and scenarios to its partners.

This situation provides the context for many GenAI-related announcements from the ServiceNow ecosystem, including the collaboration of Accenture, Nvidia, and ServiceNow to launch AI Lighthouse, a program designed to fast-track the development and adoption of enterprise GenAI capabilities. Additional announcements include Cognizant’s ambition to deliver AI-driven automation to achieve $1 billion in revenues and ServiceNow’s investment in Plat4mation to accelerate GenAI adoption at German Mittelstand businesses (small and medium or family-owned). Compared to the more marketing led and, therefore, noisier announcements, EY’s declaration comes across as much more measured and grounded. EY needs to follow up with case studies of successful deployments to avoid being labeled equally marketing led.

To accelerate the evolution of AI, double-click on integration and governance

For operations leaders, EY’s AI governance proposition provides a framework to operationalize GenAI projects and prepare their organizations for the broader and more challenging evolution of AI. As Exhibit 1 outlines, enterprise adoption of GenAI hinges on managing the intersection of cloud, data, and AI while providing effective governance focused on assuring the outcomes of GenAI-enabled services.

Exhibit 1: Enterprise adoption of GenAI is about integration and governance

Source: HFS Research 2024

However, HFS believes the hype around GenAI will calm down over the next 12 months, and we will shift to discussing the broader evolution of AI. Just as with the discussions on cloud-native transformation, we need to focus more on the outcomes and on working fundamentally differently. Thus, conversations on the operating model have to take center stage.

Operations leaders should engage with their suppliers to ensure the following initiatives to guarantee enterprise readiness of GenAI:

• AI monitoring: Blend DataOps with quality assurance to enable the monitoring of artificial intelligence (AI) applications and hold developers and users accountable for their outcomes.
• Transparency and explainability: To comply with the most looming AI regulations, you must ensure that humans understand AI systems and their decisions. Yet, most machine learning (ML) algorithms are black boxes.
• Outcome-centric governance: Pivot governance to assure the outcomes your services are contracted for. Move beyond ethics and tick-box compliance to proactively change your operating model.

The Bottom Line: To scale their GenAI deployments, operations leaders need to instill an outcome-centric governance mindset.

As the market enters a new phase of GenAI adoption, enterprise leaders need to prepare their organizations for AI’s broader evolution. Discussions have to pivot from capabilities to outcomes. Managing those outcomes with a CMDB can be a lever to instill an outcome-centric governance mindset crucial to harnessing AI’s transformational capabilities. As always, the proof will be in the pudding. We need to learn more about the challenges of this transformation journey.