Frontier AI is perishable commodity, says Kinetic IT

Organisations should treat frontier artificial intelligence as a perishable commodity rather than a lasting competitive advantage, according to Kishore Jayaram, Chief Transformation Officer at Australian technology services provider Kinetic IT.

He argued that the rapid pace of change in advanced AI models means businesses should focus less on securing access to the latest system and more on building the structures needed to adapt when models change. In his view, resilience, governance and operational discipline will matter more than any short-term lead from using a newer model.

"Organisations should start viewing frontier AI as a perishable commodity rather than a permanent source of competitive advantage. Frontier AI is becoming a perishable commodity. What's considered state-of-the-art today won't remain the benchmark for long, while access to frontier capability is becoming increasingly available through a growing number of providers. For most organisations, sustainable advantage won't come from access to the newest model. It will come from how effectively intelligence is operationalised and how quickly the organisation can adapt as the technology evolves," Jayaram said.

The comments reflect a wider business debate over whether AI value will sit primarily with model developers or with companies that can weave the technology into everyday operations without creating new vulnerabilities. As AI tools move from experimental use into critical functions, that question is becoming more immediate for executives overseeing technology, risk and service delivery.

Jayaram said the challenge is not usually the technical task of swapping one model for another. The bigger issue, he said, is that organisations build processes and dependencies around the behaviour of a specific model over time, making change harder even when alternatives are available.

"Replacing a model is rarely the difficult part. Over time, organisations build workflows, governance processes, evaluation frameworks and operating procedures around the behaviour of a particular model. The technology itself may be interchangeable, yet the operational capability built around it is often far less adaptable. That's where dependency begins to emerge," Jayaram said.

Beyond the model

Kinetic IT framed AI decision-making around three factors: capability, cost and control. In that view, capability remains the main reason many organisations adopt AI, particularly where it can support productivity, decision-making, workflow automation and service delivery. Cost includes not only licences and infrastructure, but also the expense of keeping systems flexible enough to avoid lock-in.

Control is becoming a more strategic concern, Jayaram said, because AI depends on a wider network of cloud providers, infrastructure operators, regulation and commercial agreements. That leaves organisations with limited influence over some of the conditions that shape how AI systems perform and how easily they can be changed.

The argument places AI within a more familiar corporate risk framework. Businesses already apply business continuity planning, supplier risk management and cyber resilience measures to many essential technology services. Jayaram suggested the same approach should now be applied to AI as it becomes part of operational systems rather than a stand-alone productivity tool.

That is particularly relevant in sectors where service interruption or degraded decision-making can have broader consequences. Kinetic IT works with government, defence and critical infrastructure customers, where resilience and continuity often take priority over pursuing the latest technology trend.

Leadership shift

Jayaram said executives should not assume the core task of leadership has changed, but that the speed of technological change has altered the assumptions behind many decisions. In practice, that means leaders need confidence that their organisations can revise AI choices without disrupting services or rebuilding large parts of their operating model.

"The organisations that create lasting value from AI won't necessarily be those that predict which frontier model succeeds. They will be the organisations that invest in the governance, engineering discipline and operating models that let AI capability evolve without disrupting critical services," Jayaram said.

For technology buyers, that could have implications for procurement, architecture and internal oversight. Companies may need to assess AI suppliers not only on model performance, but also on interoperability, contractual flexibility, auditability and the ease with which systems can be tested or replaced. Those considerations are likely to grow in importance as more providers offer access to advanced models and the differences between them narrow over time.

Kinetic IT, an Australian-owned technology services group with more than 1,500 employees, provides ICT services to public sector and critical infrastructure organisations. Its comments suggest the business case for AI is shifting from a race to secure the newest model to a broader effort to ensure organisations can adapt when that model is no longer the benchmark.

"The role of leadership hasn't fundamentally changed. Leaders are still making decisions under conditions of uncertainty. What is changing is the speed at which the assumptions underpinning those decisions evolve. The organisations that recognise frontier AI as a capability that will continually evolve, rather than a permanent source of competitive advantage, will be better positioned to build resilient services and realise lasting value from AI," Jayaram said.