Six checks that turn an AI First slogan into a Value First strategy

Is your "AI First" strategy accidentally helping your competition?

While you're chasing headlines, they're quietly building value.

We are now seeing the fallout from the first wave of "AI First" declarations. In hindsight, they weren't innovation. They were a predictable pattern of hype followed by retreat.

The core delusion was treating a marketing slogan as a strategy.

Take Duolingo. After its "AI First" announcement, the backlash was swift and material. Users reported cancelling subscriptions in droves¹. The company's daily active user growth slowed, and its stock price tumbled 14% in a single week in June 2025⁴ ⁵.

Or Klarna. It proudly announced AI was doing the work of 700 agents⁸. Soon after, the CEO admitted they had "gone too far," as service quality suffered and the company was forced to re-hire human agents to repair the damage⁹.

These aren't just PR blunders. They are failures of process, born from a shallow understanding of the technology.

The companies making actual progress aren't "AI First."

They are "Value First."

They use a more holistic approach. While not exhaustive, a "Value-First" approach involves asking deeper questions before committing resources. Here are a few to consider:

1. Reject the Binary (Adopt the Spectrum):Don't fall for the "automate or not" trap. Map each task to the 5-level scale of human-AI collaboration. Forcing full automation on tasks requiring human judgment guarantees failure.

2. Assume Agency & Manage Leverage:View any autonomous AI as a strategic actor, not a passive tool. This requires building new governance models to manage its inherent leverage and risk.

3. Build Compliant-by-Design Systems:Compliance is an architectural choice, not a feature. Build systems where EU AI Act and GDPR principles are embedded from day one, ensuring every AI-driven action is auditable and defensible.

1. 01Analyze Second-Order Effects: Go beyond ROI. Model the negative second-order effects before launch. What is the risk of skill decay in your team, reduced innovation, or brand damage?

5. Build Competence Through Co-Creation:Stop training passive users; cultivate active co-creators. Judging AIs capabilities isn't learned from a slide deck - it's forged by shaping the tool. Involve your team in defining goals and testing limits to build a team of owners, not just operators.

6. Prevent Strategic IP Leakage:Pasting internal data into a third-party AI is strategic IP leakage, as it can be used to train the vendor's model. Your competitive advantage becomes their feature. Are your contracts and governance strong enough to prevent this?

The goal isn't to be "AI First." The goal is to be strategically sound.The winners won't be the loudest. They will be the ones who think in systems.

This list is not exhaustive. What is a critical check you would add?