Corporate America has mandated AI adoption, and the organizational consequences are already visible

Jason Lemkin frames the shift with unusual bluntness: Corporate America issued a directive that AI must be used, and it happened fast. That top-down pressure is no longer confined to pilot programs or innovation budgets. It is showing up in who gets hired, what contracts require, and how companies are organized internally.

The clearest early signal is in agency work. Mark Rubin, whose agency operates in the e-commerce space, reports a straightforward market reality: clients will not retain firms that do not use AI tools. His agency began embedding AI clauses into client contracts over three years ago, an early move that now looks less like a differentiator and more like table stakes. The mandate has moved from optional to contractual on both sides of the client relationship.

The same pressure is restructuring how enterprise software gets adopted. Maxim Bar Kogan describes a pattern his team did not anticipate: open-source agent tools are being pulled into companies as officially sanctioned software, not through the usual IT procurement process but because CEOs are driving AI adoption from the top down. The buying motion has changed. Tools that would previously have required lengthy security reviews and vendor negotiations are being greenlit because the instruction to adopt came before the infrastructure to evaluate arrived.

Corporate America flipped flipped the switch a year ago and said, 'Thou shalt do AI in 2026.' Jason Lemkin

Organizational structure is following the same logic. Harry Stebbings cites the Coinbase CEO’s stated position that every employee must now function as an individual contributor, eliminating pure management roles. That is not a policy about AI tools specifically, but it reflects the same underlying calculation: if AI handles coordination and synthesis work, the organizational layer built around that work becomes redundant. Daniel Priestley adds that hiring screens are now probing for evidence of AI experimentation, not formal credentials in AI. Candidates who are not already using the tools are already behind.

The revenue effects are starting to appear in the numbers. Brendan Foody reports that Mercor added $300 million in net new ARR in 60 days, a figure that reflects how quickly demand can materialize when institutional mandates are in place. Harry Stebbings notes that Twilio’s net new customer count grew roughly 40% in the past year, driven largely by AI startups. Stebbings also cites Gartner data suggesting that 30 to 50% of AI dollars are coming from consolidation of existing vendor spend, which means a meaningful share of the apparent growth reflects budget reallocation rather than purely new spending. These figures, taken together, describe a market that is moving fast but not uniformly.

Patrick O’Shaughnessy offers the longer arc. AI adoption among knowledge workers sits at roughly 0.1% today, by his reckoning, and he projects it reaching 1 to 2%, then 3 to 5%, then 15% over the next four years. That trajectory matters because it is not linear. The gap between 0.1% and 15% mostly gets crossed in the later intervals, which means the current period of mandate and early adoption is setting the conditions for a much steeper climb. Rory O’Driscoll makes a structural point about why that climb is likely to sustain: those who cannot use AI-native workflows will be competed out. If leaders who can run campaigns or manage operations with AI outperform those who cannot, the 10% who have that capability today become 20%, then 40%, not because of enthusiasm but because the others are forced out.

What the evidence describes is a system in which the mandate precedes the infrastructure, the infrastructure is being built under time pressure, and the economic rewards are already accruing to early movers. The directive came first. The organizational capacity to manage what follows is a separate, still-open problem.