The Early Scale: AI Is Deployed Everywhere. It's Working Nowhere Near Expectations.

Good morning

Enterprise AI is officially past the "we're exploring it" phase and into the "why isn't it saving us money yet" phase. From the C-suite down, the question has flipped from adoption to accountability. That shift is creating real pressure for operators and marketers alike, and the numbers in this edition make it hard to look away. Let's get into it.

The Big Three

AI Is Deployed Everywhere. It's Working Nowhere Near Expectations.

Kyndryl's 2026 People Readiness Report finds AI is deployed in 57% of enterprises, yet only 11% have hit their top two goals. Workforce confidence has actually fallen alongside the deployment surge, and a separate SAP survey confirms the pattern: AI tools from Microsoft, Salesforce, and Google are accelerating in adoption, but ROI is landing in unexpected places like strategic insight rather than cost reduction. The governance gap between deployment and outcomes is widening fast.

The B2B angle: Before greenlighting another AI tool purchase, audit which of your existing deployments have a defined success metric and an owner; the 11% who are hitting goals almost certainly did that first.

Anthropic and Blackstone Launch a $1.5B Bet That AI Implementation Beats Model Quality

Anthropic and Blackstone, with Goldman Sachs backing, launched Ode, a $1.5 billion joint venture that embeds elite engineers directly inside enterprise clients to compete with OpenAI's own services push. The move signals a fundamental shift: the AI arms race is no longer about which model is smartest, it's about who can actually make it work inside a real organization. The implementation layer is the new moat.

The B2B angle: If you're evaluating AI vendors, stop scoring them on benchmark leaderboards and start asking how many dedicated implementation engineers will sit inside your org and for how long.

90% of CMOs Say AI Has Already Rewired How Buyers Find Brands

A new BCG survey revealed at Cannes Lions found nine in ten CMOs say AI is reshaping brand discovery right now, not eventually. The CMO role itself is shifting: data fluency, product alignment, and AI orchestration are now the job, not campaign management. Marketing is becoming an operating discipline, and leaders who don't make the transition are getting sidelined.

The B2B angle: B2B marketers should audit whether their brand appears accurately and favorably in AI-generated responses on tools like ChatGPT, Perplexity, and Gemini, because that's where your next buyer is researching you.

Also worth knowing

Cobot orders surged 55.6% in Q1 2026, led by life sciences and food and beverage, while automotive OEM robot demand slumped. If you're in manufacturing or facilities, the automation market is diversifying fast and pricing is becoming more accessible outside the traditional auto-industry buyer.

Global M&A hit $2.8 trillion in H1 2026, up 48% year-on-year, with AI infrastructure and corporate divestitures driving a concentrated surge. Mega-deals are dominating; mid-market operators should watch for consolidation pressure and partnership opportunities as large players reposition.

FedEx CEO Raj Subramaniam called the current supply chain disruption the biggest shift in 35 years, as tariffs and geopolitics force a full rewiring of global trade routes. Any business with cross-border supplier relationships should be stress-testing its logistics assumptions now.

Smart plays for the week

Search your company name and top three products in ChatGPT, Perplexity, and Google's AI Overviews, then document what the AI says and flag every inaccuracy or omission. With 90% of CMOs confirming AI has already reshaped brand discovery, the AI-generated description of your brand is now a first impression you're not controlling.

Pull your internal AI deployment inventory, assign a named owner and a measurable 90-day goal to every active tool, and kill or pause any deployment that has neither. Kyndryl's data showing only 11% of enterprises hitting their top AI goals points to a governance deficit, not a technology one; ownership and metrics are the fix.

When vetting AI service vendors this quarter, add a contract requirement that specifies the number of dedicated implementation engineers, their tenure commitment, and escalation paths before you sign. The Anthropic-Blackstone Ode launch is a direct signal that implementation depth, not model quality, is now the differentiator; your procurement process should reflect that.

Something to think about

A $1.5 billion joint venture built entirely around embedding engineers inside enterprises says more about where AI value is actually created than any benchmark leaderboard. The bottleneck is human and organizational, not algorithmic.

Teach me something: AI Governance Gap

The AI governance gap is the growing distance between how widely AI tools are deployed inside an organization and how well those deployments are actually managed, measured, and accountable. When a company rolls out AI tools across teams without assigning ownership, defining success metrics, or building oversight processes, the tools proliferate but outcomes stall. Kyndryl's 2026 data puts a number on it: 57% deployment rate, 11% goal attainment. The gap isn't a technology problem; it's a management architecture problem. Closing it means treating AI deployments more like new business units than new software licenses.