Every time I speak with an IT leader to understand their processes, priorities, and problems, I learn something new. Lately, a recurring theme has been the ‘true prospects’ of AI in IT and how to implement it. While folks are excited about the possibilities, they’re also cautious about giving in to the hype.
AI is far from a magic bullet for IT, even for most support use cases. This Reddit thread captures the essence of what I mean.
What I do know is this – AI can undoubtedly be an immense asset in IT, but you most certainly need the human touch to be a true people-first IT organization.
Based on my interactions and insights from our recently published study titled ‘State of AI in IT, 2024’, here are a few quick, actionable insights for IT leaders on where and how to start with AI in IT.
#1 Identify IT areas that tie back to business goals
IT and business leaders must be in sync when it comes to the adoption and the path ahead of AI in IT. Starting with a business problem or opportunity is more important than “adopting AI” just for the board or other shareholders.
IT teams can lead the initiative as they’re closer to the tech and have a background in supporting other business functions. While I’ve heard this anecdotally, our survey points in the same direction.
“The IT team was the originator of AI adoption activities in nearly two-thirds of organizations (61%), with the C-suite accounting for one-quarter (24%),” says the report.
Additionally, “Organizations where the C-suite had originated the need for AI had progressed less than those where the IT team had done this.” Moving on to the use cases, here are the top areas where AI in IT can help, according to the survey respondents.
This data can provide a helpful place to start, depending on your business context. You must, of course, map out the goal, strategy, and tactical execution and have set timelines and metrics to measure success.
Also, make sure you allocate a part of the IT budget for AI projects. According to the report, “60% of organizations allocate at least 5% of their IT budget to AI”.
#2 Build AI-specific expertise
In 2024, having a deliberate strategy for AI will be non-negotiable. Organizations must develop AI capabilities, either in-house or through strategic partnerships with product vendors, to ensure successful implementation.
Our survey found that 28% of organizations have no dedicated AI professionals in their teams. This is highly concerning.
Investing in building AI expertise in-house can ensure rapid learning, innovation, and leveraging the technology as it evolves. While hiring is one option, HRBPs and the management must invest money and effort in upskilling in-house talent to attain long-term benefits.
Vishal Gupta, CTO and CIO at Lexmark, shared an interesting perspective on the “AI team” structure. His team leans more towards a centre-of-influence approach than a centre-of-excellence approach. For instance, if an AI expert is deployed within the Finance team and works closely with them, they understand their unique challenges that can be potentially solved with AI.
With a centre-of-influence approach, the company can leverage influential individuals/internal champions who can drive significant impact. They can get deeper into specific functions, understand the problems and chart the course for successful implementation. A centre-of-excellence approach, while well-intentioned, can result in silos and slow down progress.
📘 READ MORE: How to Use ChatGPT for Effective Workplace Learning
#3 Conduct a pre-mortem discussion
Our study found that only 25% of end users are happy with their current IT support. Given that 3 out of 4 users are, in some way “not happy”, it’s high time we changed a few things things.
About 30% of end users would like 24/7 support and a combined 41% would like quicker responses and resolution. AI can be leveraged in all these areas and more.
Before you begin, though, you must conduct a pre-mortem discussion. It is always good to go around the room and get the stakeholders to answer, ‘what could go wrong’ and ‘how could this project fail’ before you start.
A few concerns shared by IT professionals in the study are ‘customer data security’ (42%) and ‘additional costs’ (39%). Surprisingly, ‘governance and compliance’ featured fairly low in the list, at 28%. There are certainly more such areas to consider.
IT teams also need a well-thought-out plan to counter these potential pitfalls. And there must be a special emphasis on change management with set pre-launch, launch, and post-launch activities. You should set milestones to track progress and keep the team motivated.
Finally, as part of the pre-mortem exercise, it is good to establish strategic AI no-go areas. For example, 41% of the respondents believe AI shouldn’t be used in ‘ethical decision-making’, according to the survey. I’m sure you will find a few more areas that you, as an organization, feel are a no-go.
You must establish clear ethical guidelines and ensure responsible and transparent AI use.
#4 Start small, but start somewhere
“Consider this an easy call to action: start where you are. Learn what your current ITSM tool can support using AI and leverage those capabilities, then grow into other areas.” - Phyllis Drucker, Author, Keynote Speaker, Service Management Thought Leader
We reached out to Phyllis for her input on the findings and returned with this gem.
This helps because it is vital to get an early-mover advantage. According to the report, 58% of organizations are in the early stages of AI adoption– either ‘planning’ (20%), ‘early exploration’ (24%), or ‘pilot projects’ (14%). In comparison, 27% of respondents have progressed past the AI ‘pilot projects’ stage to have functioning AI capabilities in IT.
The point is that a significant chunk of companies already have started to dabble in AI for IT in 2023. Many experts think that 2024 will be the year when teams will see the business impact.
While planning and having a streamlined approach is important, it is equally important to get early wins to set the tone for the rest of the implementation. Seasoned IT leader, Barry Shurkey, CIO at NTT Data, seems to agree.
“While there remains a lot we don’t fully understand about AI, including its associated risks, there are many opportunities to take advantage of moving forward in business and life,” says Barry. “Falling behind in the AI adoption race can pose significant challenges for organizations.”
Conclusion
Our study found that 75% of end users already use free AI tools like ChatGPT for their work, especially for ‘creative ideation and problem-solving’ and ‘content creation’. Thankfully, ‘generating and testing code’ featured fairly low on the list. But before this leads to a cybersecurity breach, IT needs to understand these use cases and enable the users with more secure solutions.
It is important to recognize and support end users’ familiarity with AI tools and explore ways to integrate AI seamlessly into daily IT and business workflows.
A word of caution: while deploying AI in IT, you must be vigilant about a few areas. Based on our findings, these include:
- Data security – both customer data and business data/IP
- Governance and compliance
- A legacy tech stack
- Inaccuracy and inconsistency (that might derail end-user trust)
As you navigate the AI landscape, it is essential to remember that AI isn’t a silver bullet, it can help with streamlining mundane tasks and also help quicken the pace of more complex tasks, but you do require smart humans to ensure a people-first IT support function.
We conducted the ‘State of AI in IT, 2024’ study to help IT professionals and leaders understand what their peers were doing and what they were cautious about.
Download the full report to dive into trends and equip your team for success in the AI-driven future.
Vijay Rayapati
Co-founder and CEO, Atomicwork
