Building an AI governance framework may sound daunting but it doesn’t have to be
Generative AI holds tremendous potential, yet many companies struggle to turn its promise into lasting value. The IBM Institute for Business Value (IBV) found that typical organizations aim to grow their generative AI model portfolios by 50% over three years. Yet, one in three companies pauses an AI use case after the pilot phase - indicating a struggle to move beyond experimentation. Effective governance could be the key to breaking this cycle.
In response, enterprises should stop seeing generative AI (generative AI) governance as an inhibitor to innovation. In fact, it could be the driver they’re looking for to unlock ROI. There’s no doubt that generative AI is attracting big investment, with the market for this technology estimated to double every other year. This statistic is no surprise when you consider analysis from Deloitte showing that successful digital transformation with generative AI at its center can result in up to USD 1.25 trillion in additional market cap.
But equally, there’s a gap between what organizations aspire to achieve with generative AI and the current reality. Research suggests that companies may be overstating their progress—known as ‘AI washing.’ For example, IFA Magazine reports a 75% increase in UK companies mentioning ‘AI’ in shareholder statements between 2022 and 2024. Yet, in the Nordics, only 5% of generative AI experiments are operationalized into production, revealing that many companies are investing significantly with limited real-world results.
To provide a broader view, an EU-wide study from the European Commission shows that as of 2023, only 8% of EU businesses actively use AI technologies. This data reinforces the contrast between AI ambitions and practical outcomes across Europe.
So, what’s getting in the way of the brave new world promised by generative AI?
One of the main barriers to successful generative AI adoption is a lack of joined-up thinking. It’s all too common for AI to be deployed in silos, with different teams across a company experimenting with small initiatives. Without an AI inventory organizations can end up duplicating effort and wasting resources. It’s also more likely that companies will end up pursuing use cases that shouldn’t be a priority, missing out on the bigger opportunities.
Moreover, enterprises must overcome consumer mistrust of AI to embed it successfully in their businesses. Deloitte research shows that when customers know a brand is using AI, their trust in the brand declines by a factor of 12. Companies must address this issue head-on, putting measures into place to reduce the risk of AI, bias and discrimination and that ultimately helps to foster consumer trust in the technology.
Chief data scientists have traditionally been cautious about the push for more AI governance, often viewing it as a potential obstacle to innovation. However, recent conversations within the industry and direct feedback suggest that perceptions are shifting. More and more, chief data scientists are beginning to align their views with those of chief compliance officers, who regard AI governance as crucial for ensuring safety and effectiveness. This change reflects a broader trend toward a more proactive approach to governance. As new AI regulations emerge, both data science and compliance leaders are recognizing that effective governance is key to supporting responsible AI deployment and mitigating potential risks.
Governance isn’t just about reducing risk, however. By proactively rolling out an AI governance framework, companies can accelerate their AI plans—unlocking efficiencies and gains that may have been eluding them until now. An effective framework can help organizations prioritize the right use cases, focusing their investments where they’ll generate the largest returns.
It is also crucial in building trust into AI initiatives, ensuring you train models on appropriate data sets, take sustainability targets into account and much more. By doing the qualification work upfront, companies will be able to move more of their generative AI models into production, giving them an advantage over less focused competitors.
The World Economic Forum reported that funding for generative AI reached £25.2 billion in 2023, underscoring the critical role of robust governance in attracting and retaining investment while mitigating risks.
The good news is that the frameworks for AI governance are there, such as the EU’s landmark AI Act, which is considered the world’s first comprehensive AI legislation and which will standardize regulations and places its focus on accountability and transparency, to regulate the risk – not the technology. Companies that align with these frameworks will be better positioned to manage compliance challenges, win over consumer trust and ultimately set their AI journey on a long-term path to success. Effective governance encompasses many, complex processes, such as regular bias audits to ensure fairness in AI systems, to comprehensive reporting on performance metrics like accuracy and reliability. These practices enhance transparency and foster trust among users and stakeholders, which we believe is critical to unlocking the tremendous potential of AI technology to society as a whole.
Building an AI governance framework may sound daunting but it doesn’t have to be. First things first, create a cross-function AI team. Responsibility for governing your AI initiatives shouldn’t be isolated to your risk and compliance department, as it touches all parts of the business.
Next, prioritize AI literacy, especially at the C-level, to ensure alignment across leadership. Finally, set clear goals for your AI initiatives from the outset.
These steps will put your organization in the best position to succeed. In a world where generative AI is playing a growing role in setting companies apart from the pack, it’s time for business leaders to build a bold vision with governance at its core.
You May Also Like
