The GenAI market is extremely noisy, with developers rushing to launch new platforms and capitalise on the growing demand. In fact, more than 250 LLM models have been launched in the past year alone.

For CIOs, cutting through this noise and making the right decision is challenging. Without a careful and strategic roll-out, this next-gen technology could inadvertently lead organisations to more risks than benefits.

I know AI can be a daunting challenge for CIOs tasked with driving implementation. Deploying GenAI effectively requires balancing scalability, security, and user acceptance while also avoiding the pitfalls of tech sprawl. Here's some practical advice on a successful roll-out based on my experience.

Challenges in implementing GenAI

At Jamf, we're nearly two years into using a GenAI copilot solution, and change has been a constant force. The business environment demanded rapid adaptation, particularly as we managed a hybrid workforce across 3,000 locations.

With employees spread across thousands of locations, Jamf needed a GenAI tool that could handle a wide range of tasks, from IT support to HR queries.

Security was another major concern, and the question of data privacy loomed large. I was worried that employees might be using their own AI tools without the knowledge of the IT and security teams. We needed safeguards that protected users but didn't restrict them from exploring new tools.

In the end, I decided to look for GenAI tools that were compatible with our workforce arrangement and organisational structure. There are so many SaaS platforms offering GenAI tools, and most overlap with each other. Often, I was confused about why a particular tool even existed or soon realised the extra security or usability headaches it would create.

By the end, I had to keep asking myself, "What value would this bring to Jamf?" For CIOs out there currently looking for GenAI tools, it’s important to remember that sometimes less is more. You bring no value to your workforce if you give them five or six different GenAI solutions they have to use without explaining how they all work.

It’s also critical that CIOs have a roadmap for GenAI implementation that clearly sets milestones for scalability, ensuring that as the organisation's needs evolve, the solution can adapt accordingly.

Here are five essential tips that every CIO should consider when initiating their GenAI roadmap:

1. Have a North Star

I’ve seen many organisations rush to adopt AI without a defined purpose, leading to misguided efforts and wasted resources. It essentially becomes the blind leading the blind.

Right from the start, I knew our goal was to enhance specific functions such as IT and HR, automate repetitive tasks, and improve efficiency. This meant I could measure the value and impact of each GenAI solution against our organisational objectives and, more importantly, quickly pivot if it wasn’t delivering as expected.

2. Be open to all feedback

During Jamf’s GenAI implementation, listening to end users was crucial. Not all users initially embraced the changes. Some felt threatened by the automation, particularly in areas involving code review and recommendations. Therefore, actively seeking feedback was essential.

Additionally, collaboration among different teams has never been more important with the advent of GenAI. I realised that my role as CIO required me to be in constant and regular dialogue not just with the IT team, but also with the security team and the data privacy team to ensure that GenAI systems met organisational standards.

3. Be your own harshest critic

Self-assessment is vital. As the final decision-maker, I knew I had to rigorously evaluate the performance and impact of GenAI tools.

At Jamf, a constant review process was in place to measure the value provided by the AI copilot. This included assessing whether the tool was genuinely improving efficiency and meeting security standards. When weaknesses were identified, immediate action was taken to address them.

4. Expect pushback

Resistance to change is inevitable. Even the best-planned AI implementations will encounter scepticism. When we started rolling out GenAI, it was initially well received by users. However, to our surprise, we faced pushback from engineers. There was notable discomfort with AI making recommendations or changes to their code, which disrupted their usual workflow.

We overcame this resistance by initially rolling out AI tools to pilot groups, allowing for controlled testing and refinement. This experience highlighted to me the crucial importance of how you communicate changes.

5. Understand the endgame

Every GenAI deployment must be guided by a clear endgame, not just trend-chasing. Throughout the entire deployment, I consistently asked myself: Why do we have a copilot? How will automating tasks add value?

GenAI is still relatively new, and few businesses have extensive experience with long-term deployments. However, by approaching AI with a strategic mindset and a commitment to continuous improvement, organisations can unlock its full potential and drive meaningful, sustainable growth.

Linh Lam CIO Jamf