AI “Transformation,” Part 3: The Core Issues
Two decks, one choice. Whichever path she picks, the same five core issues are waiting underneath.
If you’re just joining: our CTO, chin-deep in last year’s “transformation” priorities, has a 20-minute board slot to answer “What are we doing about AI?” In Part 2 she wrote two complete decks in one night.
Path One, The Roadmap: harvest the AI already embedded in the existing stack, request funding for a focused effort, and return in three to six months with a prioritized plan tied to the current business strategy.
Path Two, The Reset: admit the company is not structured, staffed, or funded for what is coming, and redesign the business - not the IT department - with technology at the center.
The Morning After
It’s 5:40am. The draft is due to the CEO today, and both decks are open side by side. Before she chooses, she questions her own work as objectively as she can.
The two decks - opposite in temperament, opposite in cost, opposite in career risk – fall apart in the same places. Not different strengths and weaknesses for different strategies. The same five core issues, every time.
Which is when she realizes The Roadmap and the Reset are not two distinct answers. They are two points on a spectrum. She has spent two nights choosing between paths, when what matters is how they address the core issues at hand.
Core Issue 1: Governance Built for a Slower World
Her company’s planning and governance machinery - annual funding cycles, business cases, stage gates, risk signoffs, the transformation steering committee - was built to prevent bad surprises, and it is good at its job. Every step is individually defensible. Collectively, they are calibrated for technology that changes on a five-year cycle, and they are now being asked to govern technology that changes monthly.
The Roadmap survives governance by moving at governance speed - which is both its strength and its weakness. Three to six months to produce a prioritized roadmap, another quarter to build the business cases, then the next annual planning cycle to fund them. The board’s question will be 12 months old before the first project starts - and in this market, 12 months is somebody else’s head start.
The Reset has the opposite problem: it collides with governance head-on. Unplanned investment in the middle of a budget year. Write-downs on recent capital investments. That deck will be seen as a threat, because it is. A threat to sunk cost and a threat to ego.
Last year’s investments were proposed, defended, and approved by the very people who will review this year’s ask, and CAPEX schedules lock you into yesterday’s technology, because retiring an asset ahead of its “useful life” means admitting, in the financial statements, that it was a bad investment.
What’s needed is a two-lane system: one lane that keeps governing the business exactly as it always has, and a second lane – faster timelines, looser gates, different risk appetite - for the technology that is moving faster than governance. She can think of one person in the company with the authority to build that second lane and it’s not her.
Core Issue 2: The Talent Gap
The technology organization she leads was hired for yesterday’s priorities - and correctly so. She built a team optimized to select, integrate, and manage third-party platforms, because that was the strategy, and it was a defensible strategy.
So: rebuild or upskill? The honest answer is both - set new expectations and goals for the existing team, manage performance against the new expectations, and blend new talent into the mix over time. The hard variable is the timeframe. Moving fast on people is disruptive, and it feels disloyal to a team that delivered exactly what was asked of them. Moving slow is kinder and gentler but guarantees the dial does not move within a year.
And that’s only the IT part of the gap. The bigger problem belongs to everyone else. Like the advent of the internet: basic fluency with AI will eventually be as unremarkable as using the web. Try to imagine an employee today saying that they don’t like to use the internet to do their job.
Which raises the question: whose job is it to close that gap? The board and the executive team reflexively look at the CTO - it’s “technology,” after all. But an enterprise-wide competency shift is a training and change-management effort, and it belongs to HR and the business leadership, with IT in support.
Core Issue 3: Strategic Clarity at the Board Level
“What are we doing about AI?” is not one question. One director means: are we about to be publicly embarrassed? Another means: is our expense ratio about to become uncompetitive? A third means: is our underwriting edge about to evaporate? A fourth, trained by decades of fiduciary caution, means: how do I know you won’t blow something up? Answer the wrong question brilliantly in twenty minutes and you will leave the room having reassured no one.
Strategic clarity at the board level regarding the opportunity and threats of AI as they apply to the business they oversee is a pre-requisite to a successful transformation effort.
There’s another problem: the question was addressed to the CTO. AI will reshape underwriting, claims, service, distribution, and finance - which means it cannot be the CTO’s problem to figure out alone, or every AI initiative in the company becomes “an IT project.” Ask any veteran technology leader how the business treats an IT project it didn’t ask for.
There may be a professional reflex for the CTO to want to “own AI,” but a successful CTO will foster broad ownership and act as a thought leader and facilitator to their executive peers.
Core Issue 4: Is the Business Strategy Still on Point?
Convention says prioritization should be “business-driven”: the business states its needs, and IT delivers against them. That is the right model when the business understands the technology at least as well as it understands its needs. It fails when the technology is moving faster than the business’ understanding of what is possible - you can’t ask people to prioritize capabilities they cannot yet imagine.
This moment calls for genuinely collaborative prioritization: technologists who understand the business well enough, and business leaders curious enough to engage with proposals that don’t map to anything they’ve seen before.
The problem is when every plan in the company is required to “align to the current business strategy.” But what if AI calls the current strategy itself into question? Alignment to the current strategy is a virtue only if the strategy survives contact with the new reality.
Core Issue 5: Shaky Foundations
The company’s processes and workflows were designed for a human-intensive business. The handoffs, backlogs, referral thresholds, and audits all exist because humans have limits and make human mistakes. Apply AI to a workflow designed around existing constraints and you mostly get the old constraints executed incrementally faster.
And underneath these processes sit bad data: scattered across systems that don’t share a common view of the customer or shared definitions, managed in myriad spreadsheets, reconciled by hand, and sitting in shared folders. Every ambitious AI use case is blunted by this reality.
For certain companies, the culture itself may be a shaky foundation to embrace AI. Many insurance company cultures are risk averse, and do not embrace experimentation or tolerate failure. The disruptive and rapidly evolving nature of AI requires a level of risk tolerance to reap the benefits.
“The Roadmap” mostly ignores this problem. It relies on vendor AI features built on top of a shaky foundation, which is why its use cases will likely be incremental, and its benefits will be modest. “The Reset” rebuilds the foundation, which is why its price tag could be much larger than anyone is prepared for.
The CTO Dilemma
How many of these issues does the CTO directly control? A new governance model requires the CEO. Enterprise-wide upskilling belongs to HR and the business. Strategic clarity at the board. Pressure-testing business strategy against the new AI reality requires the P&L leaders. Fixing the foundations requires multi-year, cross-functional commitment.
The dilemma for the CTO is that they are being asked a question that is an enterprise-wide issue, not only a technology issue. The board and CEO expect answers, but there are many more questions.
It’s 6:55am. The draft deck is due today. She grabs her iPhone and types a simple message to her CEO:
“We need to talk…”
What she asks for in that meeting, and how the CEO responds, will be covered in Part 4.