Anthropic Just Bought the Big 4. OpenAI Just Bought Banking.

PwC and Anthropic expanded their alliance with a commitment to certify 30,000 US PwC professionals on Claude Code, then extend access to the firm's full 364,000-employee workforce across 136 countries. PwC also launched a standalone “Office of the CFO” business unit built on Claude, targeting banking, insurance, and healthcare. The joint pitch: tackle the $2 trillion in enterprise tech debt sitting in legacy systems, data silos, and manual processes. Production results already on the board — insurance underwriting cycles compressed from 10 weeks to 10 days, cybersecurity incident response from hours to minutes, mainframe modernization tracking on time and under budget.

Last Thursday's edition framed OpenAI's Deployment Company launch as the moment the platform race got serious. PwC × Anthropic is the same thesis, validated 96 hours later by the biggest professional services firm in the world. And it's not alone: same week, OpenAI scaled BBVA's ChatGPT Enterprise rollout 10x to all 120,000 employees globally (see Quick Hit #1) and announced a Deutsche Telekom partnership aimed at “millions across Europe” (Quick Hit #2). Two frontier labs, two Fortune-50 distribution deals, two days. The deployment layer isn't the next phase of the platform race — it IS the platform race.

For SMBs the read is the same as last week, now louder: when 30,000 PwC consultants and 120,000 BBVA employees are running Claude and ChatGPT in production, the verticalized SaaS layer that follows arrives faster than the historical 9–18 month trickle-down. The Office-of-the-CFO playbook PwC just launched is the template every vertical will copy. The signal to watch isn't which Big 4 firm signs next — it's which mid-market practice firm packages the same workflows as $500-a-seat SaaS first.

Delivery improvement PwC reported on Claude rollout — insurance underwriting cycles compressed from 10 weeks to 10 days, cybersecurity incident response hours to minutes. Last week's Deployment Company thesis just got validated in production at the biggest professional services firm in the world.

Source: Anthropic / PwC · May 14, 2026

BBVA and OpenAI expanded their collaboration into a multi-year strategic AI transformation program that puts ChatGPT Enterprise into the hands of all 120,000 global BBVA employees — a 10x increase from the bank's prior 12,000-seat deployment. BBVA is the Spanish global bank PwC's healthcare playbook will model: a tier-1 institution running AI on every desk, every customer interaction, every back-office function. The OpenAI version of the deployment thesis just landed at the same scale as Anthropic's PwC one — except instead of training consultants, it trained a bank.

OpenAI announced a new collaboration with Deutsche Telekom — Europe's largest telecom carrier — to bring advanced AI capabilities to millions of customers and employees. The play is distribution: Telekom's 250M+ wireless customers, 50M+ fixed-line subscribers, and 200,000 employees give OpenAI an EU consumer + enterprise channel that bypasses the platform-by-platform buy-in cycle. Coming the same week as BBVA's scale-up, this is OpenAI executing a European pincer: financial-services adoption inside, telecom distribution outside. Anthropic's PwC deal is the consulting channel. This is the carrier channel.

Anthropic split Claude subscriptions again, emailing Max 20x subscribers on May 13 that effective June 15, the Claude Agent SDK, claude -p, GitHub Actions, and any third-party tool authenticating through a Claude subscription move onto a separate metered credit pool at standard API list prices. Pro accounts get $20/month, Max 5x gets $100, Max 20x gets $200, Team gets $100/seat, Enterprise gets $200/seat. Unused credits don't roll. Interactive Claude Code, Cowork, and chat stay on existing subscription limits. The math: an effective 12-175x price increase for heavy agent workloads. Operationally — if you're running Claude in production via the Agent SDK, your June 15 budget just changed. Two weeks to model it.

AI video pioneer Runway closed a new round at $5.3B post-money and added $40M in annual recurring revenue in Q2 2026 alone, according to one of its founders. The strategic frame in the TechCrunch piece is the part to read: Runway is no longer positioning as the AI-for-filmmakers play. It's positioning as a multimodal frontier lab — text, image, video, world models — competing directly with Google's Veo line. For enterprise marketers and content teams, the read is that AI video moved past “proof of concept” faster than text did, and the vendor map is consolidating around three names instead of thirty. Procurement decisions on video AI in H2 2026 lock in for years.

Z.ai's GLM-5.1, MiniMax M2.7, Moonshot's Kimi K2.6, and DeepSeek V4 all shipped inside a 12-day window last month — each one benchmarking near Western frontier on agentic coding at a fraction of the inference cost. Kimi K2.6 is the first open-weight model to beat GPT-5.4 on SWE-Bench Pro (58.6 vs 57.7); Chinese frontier is now 15–30× cheaper than international peers for comparable workloads. The DEV piece sorts the real numbers from the marketing math — and explains why procurement's first question for any Western enterprise is about to be about jurisdiction, not capability.

I appreciate your support.