The Urgency Curve

Bain argues the window for establishing foundational AI capabilities is closing. Get in now or get left behind forever. I've heard this before. Cloud computing. Digital transformation. Mobile-first. Every platform shift comes with the same warning: move fast or die.

Sometimes it's true. Early movers in cloud infrastructure did build durable advantages. But early movers also paid the pioneer tax. They absorbed higher costs, made architectural mistakes, and often got leapfrogged by fast followers who adopted more mature, commoditised solutions at lower risk.

The question isn't whether AI matters. It does. The question is whether your industry structure, competitive dynamics, and capital position actually require you to move right now, or whether a measured fast-follow strategy makes more sense. In sectors with rapid product cycles and digital-native competitors, delay is costly. In asset-heavy or regulated industries, rushing in without proper governance is reckless. Bain's framing assumes one timeline fits all. It doesn't.

The Real Constraint

Bain's most useful insight is organisational, not technical. AI strategy can't stay confined to IT. Scaling requires CEO sponsorship, board-level oversight, and changes to operating models, talent strategies, and capital allocation. This is where most enterprises actually fail. But that takes time. It takes iteration. And it takes the flexibility to change direction when your initial assumptions collide with reality.

The Uncertainty Curve

Most consulting engagements work backwards. They ask you to commit the most money when you know the least. Big contracts. Long timelines. Locked-in scope before you've tested anything in the real world. You wouldn't do that with a house. But enterprises do it with AI initiatives all the time.

The smarter approach is the opposite. Spend the least when uncertainty is highest. Run a small pilot. Test an idea. Learn something first. Then increase investment as certainty grows. This isn't radical thinking. It's common sense. But it's also the opposite of how traditional consulting firms are incentivised. They want large orders and long-term commitments upfront. That's how they derisk their own business, not yours.

What Actually Works

Winners build reusable GPTs and API-powered assistants. BBVA regularly uses more than 4,000 custom GPTs. Winners have clear executive mandates, dedicated resources, and space for experimentation. They codify institutional knowledge into machine-readable formats and run continuous evaluations against real-world outcomes. None of this requires a Code Red mentality. It requires discipline, clarity, and the flexibility to iterate as you learn.

The Trust Problem

Here's what Bain's report doesn't address: trust is collapsing as adoption scales. In 2025, 83% of AI leaders say they feel major or extreme concern about generative AI. That's an eightfold increase in just two years.

The worries are familiar but intensifying. Implementation costs balloon faster than expected. Data security questions grow. Outputs remain unreliable. Decision-making lacks transparency. Only 27% of organisations express trust in fully autonomous AI agents. That's down from 43% one year earlier. Fewer than 20% of organisations report having mature data readiness. Over 80% lack mature AI infrastructure. You can't scale what you don't trust. And you can't build trust by moving faster.

The Governance Gap

Speed without governance introduces material exposure. As AI systems gain broader access to enterprise data, privacy and cybersecurity risks expand. Model hallucination can propagate incorrect outputs at scale if guardrails are insufficient. Bias embedded in training data produces discriminatory decisions in regulated contexts like lending, hiring, or insurance.

Scaling AI also increases financial exposure. Infrastructure costs, licensing, and specialised talent accumulate rapidly. If initiatives aren't tightly aligned to core value drivers, enterprises risk substantial stranded capital. A disciplined governance architecture should include model validation, human-in-the-loop controls for high-stakes decisions, clear data lineage, and board-level visibility into AI risk posture. That's not something you bolt on after the fact. It's something you build from the start.

Is It Code Red?

Bain's stature and access to executive sentiment lend weight to their warning. But leaders should distinguish between urgency as a catalyst and inevitability as a conclusion. AI has moved beyond novelty. For large enterprises, it now intersects directly with strategy, risk, and capital allocation.

But that doesn't mean every organisation faces the same timeline. It doesn't mean every pilot should be scaled. And it doesn't mean late adopters are destined for permanent disadvantage. What's clear is that AI transformation starts and succeeds in the C-suite. Grassroots experimentation sparks innovation and cultural momentum. But it doesn't self-organise into enterprise-wide impact.

Without clear direction from the top, efforts remain fragmented, siloed, and shallow. The organisations making real progress combine ambition with discipline. They scale where value is demonstrable. They govern where risk is material. And they align technology investments tightly with core business outcomes.

That's not a Code Red moment. That's just good business.