AI Featured

Strategy

3 lessons risk leaders have learned on their AI journey

by Sterling Thompson

3 min read

May 1, 2026

The risks associated with the adoption of Generative AI are a primary concern for senior risk leaders. In response to member demand, we conducted an exclusive benchmark to explore how organisations are navigating this landscape; identifying the common roadblocks encountered and the strategies used to overcome them.

The benchmark highlighted 3 core challenges risk leaders face when balancing managing risks with the pursuit of opportunity:

1

Overcoming the velocity gap – innovation vs control

2

Remembering the human element – the ‘converted’ vs ‘the opposed’

3

Building technical skills – threats vs opportunities

In this insight blog, we explore each of these challenges in turn, examining what they mean, and the lessons risk leaders have learned in their drive to overcome them.

1. Overcoming the velocity gap

A primary source of friction risk leaders identified is the tension between innovation and control, specifically around what data the AI has access to, and output reliability.

On one hand, a “business rush” around AI, or fears of “missing opportunities” due to slow pace of adoption, reveal a collective anxiety about falling behind competitors. This could be seen to make the case for quicker AI adoption, when balanced with controls, to ensure this fear does not become a reality.

Simultaneously, leaders are struggling with “visibility” around AI tools. Employees are frequently and proactively seeking out AI tools, and using them before formal guardrails are established.

While risk appetite for AI adoption varies by organisation, striking the correct balance between speedy implementation and effective guardrails is crucial.

As one risk leader suggests, “guardrails don’t need to be perfect”. The priority is establishing boundaries quickly, in step with the speed of adoption.

How are your risk peers using AI to turbocharge risk management?

Join our virtual talk for a real-world look at how risk leaders are embedding AI into their day-to-day work.

Register now

2. Remembering the human element

Risk leaders tell us that AI is often not just a technical challenge for an organisation, it is a divisive, cultural issue. One of our members identified a clear split developing within their organisation, between the “converted” (early adopters) and the “opposed”. The reluctance of the opposed is driven by two factors:

A lack of understanding about the opportunities AI creates
Concerns about job security

While automation and machine learning will inevitably reshape certain roles, the Generative AI tools currently used in corporate contexts remain imperfect. Risk leaders in our network are clear that “you can’t completely remove the person” from key processes.

Accountability must be ensured and biases must be managed. By educating the workforce and providing training on responsible AI use, organisations can alleviate these cultural anxieties and move towards a more cohesive AI adoption strategy.

Emerging Risk Insights to action blog images square

Using AI to elevate risk management

Download 3 detailed use cases from practising risk leaders in our network

Find out more

3. Building technical skills

Another significant challenge for risk leaders is the pace of technological change. AI development is outstripping the rate of employee upskilling.

In some instances, teams responsible for governing AI lack the technical skillset themselves, leaving them blind to both threats and opportunities. As one member noted: "How can you build guardrails if you’re not fully up-to-date on capabilities?"

Leaders are also hyper-aware that governance is only as effective as the underlying datasets, which can be prone to bias and inconsistency.

To overcome these barriers, companies are increasingly utilising “closed environments” – secure, controlled settings that provide a safe space for experimentation. Testing AI tools in a sandbox environment serves a dual purpose: it identifies technical inconsistencies and allows employees to build vital AI skills in a low-stakes, trial-and-error setting.

Member case study

One of our members recently briefed their Risk Engagement Manager on the specific intelligence they needed to generate for their business. By using an LLM in conjunction with our Horizon Scanner, we worked with our member to produce a a risk taxonomy table.

This provided a clear breakdown of emerging risk scenarios, the business units most likely to be affected, and a cascade analysis of how these risks interconnect. The output also included a set of practical, targeted questions the member could take directly into internal workshops or board meetings to spark informed discussion.

Read more

What's next?

Nearly a third of the companies we recently surveyed have actively adopted AI. It’s clear that the technology is no longer just a talking point for organisations, but a reality.

On 19th May, we’ll present a real-world look at how risk leaders are moving beyond experimentation and starting to embed AI into their day-to-day work at the latest RLN Talk. Find out more and register for your place here.

Alternatively, if you’d like to know more about the benchmark this blog draws from, talk to our team to arrange a walk-through of the findings.

walkthrough icon

Find out more about the benchmark that inspired this blog

Request a walkthrough

Share this

Previous post

← 4 risk responses to ongoing conflict in the Middle East

Get new posts by email