Still figuring out how to use AI for your organisation? You're not alone

Most organisations are firmly on the ground floor of AI adoption, with many lacking the basics needed for sustainable — much less meaningful — AI implementations. That’s a near-universal sentiment shared by hundreds of IT pros in the latest SolarWinds IT Trends Report. Over 88% of respondents have begun charting their organisation’s path to an AI-centric future, but there are concerns over the lack of infrastructure and fail-safes needed to turn AI from a pipe dream into a lean, mean, productive machine.

Interestingly, almost half of all respondents considered AI investment insufficient (56%) and too slow (46%) within their organisations. The key takeaway from all this? Business AI adoption is in its infancy, and IT pros shouldn’t be swept up by the wave of urgency that seemingly surfaces whenever AI is mentioned. Time should instead be spent devising a value-adding AI strategy for their organisation — and a good starting point would be the concerns and considerations highlighted by their peers below.

1. Lead with pragmatic and measurable use cases

Like any other technology before it, the most beneficial implementation path for AI is realistic, measurable and relevant for the business. Thankfully, most IT pros agree, with 46% citing increased efficiency as their primary driver for adopting AI. The hope is that AI can prove a powerful tool and ally against the many issues plaguing IT today — from increasingly complex and costly IT environments to cybersecurity concerns, shrinking talent pools and more.

ITOps has emerged as the ideal testbed for early AI implementations. Dubbed ‘AIOps’, the integration of AI with the most repetitive and time-consuming areas of IT has produced encouraging results: over 31% of IT professionals recognise the value of AIOps in automating repetitive tasks, and 27% find that AIOps simplifies troubleshooting with real-time insights and predictive analytics.

Emboldened by these results, optimistic IT pros are incorporating AIOps into their broader strategies. Over 38% have already implemented AIOps, and 49% plan to do so in the future. For those still developing their approach, focusing on measurable applications like AIOps can provide the tangible results needed to justify further investment and gain top-down support for broader AI adoption.

2. Develop internal AI guidelines and frameworks

Unsurprisingly, optimism for AI’s potential is neatly counterbalanced by growing concern over the risks it poses to security and privacy. Nearly half (48%) of surveyed IT professionals worry about AI using personal data in unauthorised or unethical ways, and 43% are concerned about its security implications. This concern has persisted ever since ChatGPT captured the imaginations of the public and business executives alike.

What’s concerning is that IT pros plan to address these challenges with a ‘wait and see’ approach. While 64% of respondents have developed internal frameworks to address AI concerns, most are relying on government intervention to forge the path ahead for privacy and security. This sentiment is especially strong when it comes to data quality — with over 55% believing regulation is essential to combat misinformation from biased or incorrect data, and 50% thinking tighter rules will enhance transparency in AI model training.

Governments are responsible for dictating the reach and impact of AI, but they are also infamously slow to respond to technological change. Meanwhile, AI development continues at a breakneck pace, creating a dilemma for IT professionals: wait for slow but inevitable regulations and risk falling behind, or establish internal AI usage rules that may quickly become outdated.

I advocate for the latter approach. With their deep understanding of AI and its risks, IT professionals are well-suited to establish the ground rules for responsible AI use across the organisation. By leveraging foundational frameworks like SolarWinds’ AI by Design principles, they can build safeguards for everything AI-related, from customer data usage to AI insights for decision-making and even the unsanctioned deployment of shadow AI — a brand new nightmare for IT teams.

These rules and frameworks can help ensure the ethical and mindful use of AI as it’s deployed, while preparing the business to fluidly adapt to future regulations as they are introduced and enforced.

3. Consider nurturing an accountable AI culture

In the same vein, IT pros also have much to contribute to building an accountable AI culture. Only 54% are somewhat comfortable about involving AI in the decision-making process, even with human oversight — this caution is an essential counterweight to the open-eyed excitement and blind trust that other users might display toward AI’s remarkable, almost magical, analysis and forecasting capabilities.

IT pros should take this opportunity to establish strong principles of accountability — supported by the rules and frameworks above — when building AI tools or implementing AI solutions across various business functions. For starters, keep the need for human oversight front and centre; a real person must assess AI insights and recommendations for accuracy, biases and ethical concerns before it’s included in the decision-making process. Secondly, mechanisms should be established to evaluate the accuracy and consistency of AI outputs over time. This will allow organisations to keep AI ‘honest’ to its intended operating parameters.

In the grand scheme of things, business adoption of AI has barely begun, and nobody knows what the next few years might bring. However, the foundations above can help ensure that planned AI implementations have a greater chance of succeeding and — who knows — evolving into areas of demonstrable impact and value for IT in the future.

Top image credit: iStock.com/cherdchai chawienghong