Build strategic artificial intelligence, but buy for lower research and development risk

By 2025, 10% of global IT spending will be on artificial intelligence (AI), according to the Let’s get real about AI report from OC&C. The research found that 85% of executives believe new AI-powered entrants will take market dominance in their sector.

James Walker, partner at OC&C, said: “The current adopters of AI are outperforming those yet to invest in the technology. Therefore, it’s essential that businesses find out how best to deploy AI and use it as a transformative catalyst.  

“By working out the pivot points where AI can make the biggest impact, businesses can plan for tomorrow and make sure they’re not left behind by competitors.”

Looking at whether to build or buy in AI capabilities, the research found that buying modular services is better for most businesses seeking to optimise their existing operations.

“This allows companies to access best-of-breed specialists across a broad array of fields without incurring R&D [research and development] risk,” OC&C noted.

But if a business wants to use AI as an integral part of its proposition, OC&C recommended that it should build – or acquire – the full required suite of AI capabilities, which allows the company to maintain its differentiation from the competition.

From an organisational perspective, OC&C did not anticipate the emergence of a chief AI officer. Instead, its research found that in the early stages of AI maturity, business function leaders are driving use cases on their own, meaning AI is close to the benefit cases proposed to boards.

However, with time there will likely be a duplication of skillsets across multiple units, according to OC&C.

“Our expectation is that in the medium-term future, AI initiatives will reach such a critical mass that it will be critical to organise around a hub-and-spoke model – a centralised AI team serving multiple business functions with local commercial expertise close to the application. It is better to let business function leaders own AI for the moment and keep AI close to the business side,” the report noted.

While over time, a critical mass of AI implementations might mean it will be better to centralise AI activities, OC&C recommended that an organisation’s existing analytics expertise should be involved in any reorganisation of AI.