Published on the 12/02/2026 | Written by Heather Wright
The only strategy is the business strategy…
How’s your MVM? With the AI hype continuing, vendors promising transformation and boards relentlessly pushing for AI progress, Forresters’ Frederic Giron is warning that without a baseline of minimum viable maturity across core IT capabilities, most Australian and New Zealand organisations will fail to turn AI into profit.
Giron, Forrester VP, senior research director, told iStart the real determinant of AI success isn’t the pace of AI innovation, it’s whether organisations have reached the minimum viable maturity required to safely and profitably scale AI. And right now, many haven’t.
“There is a lot of excitement, a lot of activity, but usually not clearly linked to a strategic business objective.”
He says CIOs must ‘take back control of the narrative’ and shift attention from artificial general intelligence timelines to whether their IT is ready to deploy AI safely and at scale.
“There are two important factors that will influence IT organisations in 2026; only one of them is under the CIOs control,” Giron says. “AI will continue to progress, with or without AGI… but what’s really important is to decide how ready you want to be from an IT organisation perspective to deploy these AI capabilities.”
Forrester’s newly released CIO’s Guide to AI Readiness formalises that readiness test, arguing that returns emerge only where advancing AI tech meets sufficient IT capability maturity – not from faster models alone. It maps readiness across seven domains: Governance, security and risk, application and product delivery, data and information, architecture and portfolio, IT workforce, and infrastructure platforms and operations, and urges CIOs to assess current maturity, identify gaps and build a phased roadmap before scaling high-stakes AI.
Giron says the concept of minimum viable maturity is a ‘health check’ that helps CIOs avoid scaling on weak foundations. “Let’s look at your IT organisation and take stock of what is the minimum viable maturity that you need to have in order to safely and profitably scale these capabilities across the business.”
Productivity doesn’t pay the GPU bill
Giron warns that today most organisations are counting hours saved from AI use, rather than redesigning work, in a move that leaves no impact on margin, revenue or customer outcomes.
“Most companies are counting the number of hours that they’re saving their employees, which is a good productivity metric to track, but it’s not going to pay your GPU bills.
“If the productivity is just making your employees complete their tasks faster without any kind of impact on how the business is realising value… then you know it is useless.”
Companies need to understand ‘at the very minimum’ how the technologies can be applied in processes that will accelerate cycle times and increase meaningful customer interactions.
The report echoes that stance, cautioning against ‘impact disappointment’ – pilots that dazzle in isolation, but collapse at scale due to immature governance, data quality or operational outcomes. It cites the example of a retailer deploying an onboarding agent which ultimately creates fraud exposure in production because security controls designed for human workflows can’t govern autonomous agent decisions or an inventory optimisation agent making recommendations which seem reasonable individually but compound into supply chain disruptions due to data quality issues that remained invisible until the system operated at scale.
First fix: Strategy alignment
Asked where maturity most often lags, Giron is unequivocal: Strategy and strategy alignment. It’s the area where companies need to start, he says.
Giron cites Telstra’s approach – Dayle Stevens, Telstra data and AI executive, has said Telstra doesn’t have an AI strategy: “The only strategy is the business strategy”. Every AI investment is reviewed against clearly stated strategic business objectives, be that around improving customer experience, driving operational excellence or other areas.
“I see very few organisations that have this clarity. There is a lot of excitement, a lot of activity, but usually not clearly linked to a strategic business objective,” Giron says.
“That would be the first area to look at: How do you ensure you have a portfolio of AI initiatives that are clearly aligned to what the business is trying to achieve?”
Governance is the next urgent fix. With enthusiasm building around agentic tools such as OpenClaw and coding tools, Giron warns of the potential for significant fails, with the potential for big business impacts. He cites a hypothetical example where an employee deploys an agentic capability to automatically generate contracts and send them to prospect. Agents, he notes, can hallucinate, and without oversight could push out contracts on terms legal would never approve.
He urges firms to adopt ‘least agency’ – a zero-trust-like principle that would constrain what agents can do.
The report provides the scaffolding for that, extending security and risk practices via what Forrester calls its AEGIS framework, and calling for continuous controls, agent inventories, identification of models that treat agents as first-class identities and policy-as-code guardrails around high-risk actions.
“There is going to be energy, there’s going to be activities. Let’s make sure that these activities are happening in an controlled environment where people are not going to be doing stupid things that will damage the brand or impact the business [negatively].”
Don’t boil the ocean on data
For many, data remains a sticking point, but Giron argues it shouldn’t stall progress. “You could spend years and millions of dollars in investments to break down data silos and get all the data ready from a quality, accessibility and governance perspective. Is that required? Well, that depends on what you want the data to do for you.”
While data will be important – and data quality and master data management are extremely important – Giron advocates a more pragmatic approach: Align to business strategy, target specific outcomes and use AI to help accelerate data quality and product creation.
“You don’t need to boil the ocean. You can get started with much more focused initiatives aligned to the business strategy, then use AI to help you with data quality issues and the creation of data capabilities and go from there.”
The CIOs Guide to AI Readiness data capabilities determine the quality of analytics and AI more than model sophistication and says quality AI deployments require data platforms architected for AI workloads (such as lakehouses, vector stores, knowledge graphs), but reiterates to stage investments by risk and value, not as an upfront precondition to all experimentation.
The board conversation: Harness energy, fund foundations
Giron has a ‘sobering’ message for boards: The excitement is warranted, but so is investment in core IT capabilities. “That is the challenge and opportunity for CIOs – to harness this energy, but to do it in a way that also brings clarity in terms of the investments required to make sure these AI initiatives are going to scale in a safe and productive manner.”
He dubs it a ‘gymnastic’ communication challenges CIOs need to master to ‘surf on this wave of excitement’ without outrunning foundations.
Forrester’s report calls on CIOs to proactively educate the board and execs on realistic timelines, risk trade-offs and progress via concrete governance channels.
Its overall roadmap distils into three steps: Assess capability readiness, use the AI readiness matrix – which includes four AI readiness scenarios from sustained transformation to the ‘nightmare scenario’ of ‘rainbow chasing’ – to reframe the board conversation and gain commitment, and build a phased roadmap that enables near-term innovation while maturing governance, data, security and workforce capabilities for higher-stakes deployments.
For A/NZ, Giron cautions that success will be driven by leadership clarity. “It really depends on the people… whether the CEO has the vision of what AI means for the business strategy.” The lesson from digital transformation still holds: Share accountability across the c-suite, collaboration with the CIO and relentless alignment. “It’s a leadership issue, it’s a people issue, it’s not a technology issue.”
His closing message to CIOs and IT leaders? The opportunity is ‘massive’, but the wins go to those who channel the hype into capability-building. “AI is going to transform the way we work. But the opportunity for CIOs is to try and guide this energy towards making the investments that will prepare the organisation to thrive in this area.”
The first question then? How’s your minimum viable maturity?
