Get tomorrow's brief in your inbox
Today: Mercedes-Benz rolled out n8n as its global automation platform to scale AI across 164,000 employees. GM laid off 600 IT workers to hire AI-focused engineers, signaling a deliberate skills swap. Thinking Machines (Mira Murati's startup) revealed interaction models that respond in 0.40 seconds, mimicking natural conversation.
Enterprise AI adoption is shifting from proof-of-concept to production, and infrastructure is the bottleneck. Mercedes-Benz deployed n8n, a low-code automation platform, as its global AI orchestration layer. The deployment spans R&D, production, sales, financial services, HR, and IT operations. What matters here is not the tool itself but the strategy: Mercedes needed a platform that could let non-engineers build and deploy AI-powered workflows without waiting for developers. They chose n8n because it runs on their own infrastructure (data sovereignty under GDPR), connects both modern SaaS and legacy in-house systems, and gives "Makers" (employees who aren't deep engineers) the ability to orchestrate AI agents in production.
This is what scaling AI actually looks like. Mercedes ran a company-wide hackathon where employees across business units developed AI workflows using n8n. The best ideas moved into production. Customer support workflows now handle recurring issues autonomously while routing complex cases to humans. Sales teams use orchestrated AI agents across pre-sales and advisory systems. IT operations automate log analysis and anomaly detection. The key takeaway: if you're planning to scale AI beyond a handful of pilot projects, you need infrastructure that lets domain experts (not just engineers) build workflows that connect to your existing systems.
If you run a service business with complex customer workflows or IT operations with manual processes, this is your blueprint. Start with a platform that can connect your existing tools (CRM, ticketing, knowledge bases) and let your team build workflows without starting from scratch every time. Source
Zapier MCP (Model Context Protocol) makes your app connections portable across AI tools. If you switch from ChatGPT to Claude or from Cursor to another coding agent, your Zapier workflows and app connections stay intact. Zapier connects to over 9,000 apps, and those connections live at the platform level, not inside any particular AI tool. This solves a real problem: every time you migrate to a new AI tool, you typically have to re-authenticate apps, rewrite agent instructions, and rebuild governance guardrails. With Zapier MCP, any AI tool that supports the Model Context Protocol can tap into the same pool of apps and actions. Your custom GPTs, memory, and connectors don't leave ChatGPT today. Your Claude Projects don't leave Claude. Zapier sits below the harness level, so your setup travels with you. Source
Quo (formerly OpenPhone) automation via Zapier. Quo is a shared business phone system for teams managing calls, texts, and customer contacts. Zapier now integrates with Quo to automatically sync contact records between your CRM and phone system, send instant follow-up messages when leads submit forms, and log every call and text to spreadsheets or databases for reporting. If you're manually copying customer interactions from your phone system into your CRM or lead management tool, this eliminates that process. Free tier available via Zapier's free plan (100 tasks/month). Source
GM laid off 600 IT workers (10% of its IT department) to hire engineers with AI-native skills. This is not a headcount reduction. GM is still hiring for its IT department, but for different capabilities: AI-native development, data engineering and analytics, cloud-based engineering, agent and model development, prompt engineering, and new AI workflows. The company is looking for people who know how to build with AI from the ground up (designing systems, training models, engineering pipelines), not just use AI as a productivity tool. GM's CTO Sterling Anderson (co-founder of Aurora, the autonomous trucking startup) joined in May 2025 and has been consolidating GM's disparate technology businesses into one organization. The company hired Behrad Toghi (formerly Apple) as AI lead in October and Rashed Haq (formerly Cruise) as VP of autonomous vehicles.
For the industry, this is a signal of what enterprise AI adoption actually looks like: not adding AI tools on top of existing teams, but deliberately rebuilding the workforce from the ground up. If you're running an IT department or service business, the specific capabilities GM is hiring for (agent development, model engineering, AI-native workflows) point directly at where demand is heading. You don't need to lay people off to make this shift, but you do need to upskill existing staff or hire for these capabilities. Source
OpenAI launched the Development Company and acquired Tomoro to accelerate enterprise AI adoption. The new business unit is a partnership with 19 investment and consultancy firms (Bain, Goldman Sachs, SoftBank) and brings 150 engineers specializing in deploying frontier AI models. These "forward-deployed engineers" will sit with organizations, understand workflows, and help connect back-office applications to AI models. OpenAI's Chief Revenue Officer Denise Dresser says enterprise adoption is "at a tipping point." Enterprise now makes up more than 40% of OpenAI's revenue, and the company expects that portion to reach parity with consumer by the end of 2026. Source
OpenAI launched Daybreak, a security-focused AI initiative to detect and patch vulnerabilities before attackers find them. Daybreak uses the Codex Security AI agent (launched in March) to create threat models based on an organization's code, validate likely vulnerabilities, and automate detection of high-risk ones. This is OpenAI's answer to Anthropic's Claude Mythos (announced in April), which Anthropic claimed was too dangerous to publicly release and only shared privately as part of Project Glasswing. Daybreak combines GPT-5.5 with Trusted Access for Cyber, GPT-5.5-Cyber, and Codex. If you're running security assessments or vulnerability scanning, this is a tool worth watching when it becomes available. Source
Thinking Machines (Mira Murati's AI startup) announced "interaction models" that respond in 0.40 seconds, roughly the speed of natural conversation. The model processes your input and generates a response at the same time (called "full duplex"), so it's more like a phone call than a text chain. Right now, every AI model works the same way: you talk, it listens. It responds, you listen. Thinking Machines is trying to change that. This is a research preview, not a product. A "limited research preview" is coming in the next few months, with a wider release later this year. The benchmarks are impressive, but we won't know if the real-world experience lives up to the technical claims until people can actually use it. Source Source 2
GitLab announced workforce restructuring focused on the "agentic era." The company is reducing the number of countries where it operates by up to 30%, flattening the organization by removing up to three layers of management, and re-organizing R&D to create roughly 60 smaller, more empowered teams with end-to-end ownership. GitLab's stock price dropped from ~$52 a year ago to ~$26 today, corresponding to uncertainty about the company's growth as agentic engineering impacts its core market. GitLab's bet: "As the cost of producing software collapses, demand for it will expand." If your business depends on software engineering growing as a field, you have a strong incentive to believe that agents will have that effect. Source
ChatGPT adoption broadened in Q1 2026 across age, gender, and geography. Users with typically feminine names now account for over half of users (for whom gender can be inferred). Messages from users over 35 gained share this quarter. The 10 fastest-rising countries by messages per capita came from Latin America, the Caribbean, Asia-Pacific, and Africa. Within work-related usage on consumer plans, creating written and visual materials continued to lead, but more specialized tasks (content creation, health-related documentation, information retrieval) became more popular. The data suggests ChatGPT is becoming a mainstream tool used by a broader mix of people, in more countries, and for increasingly recurring tasks. Source
If you're paying developers to build custom API integrations, stop. Use Zapier or n8n to connect your tools without writing code. If you're switching AI tools every few months, set up your workflows on Zapier MCP so your app connections travel with you. If you're manually copying customer data between systems, automate it with triggers and actions.
If you run an IT department or service business, audit your team's AI-native skills. Can your engineers design AI systems, train models, and build data pipelines? Or are they just using AI tools for productivity? GM's hiring priorities (agent development, model engineering, AI-native workflows) are a roadmap for where enterprise demand is heading. Upskill existing staff or hire for these capabilities now.
Enterprise AI adoption is moving from isolated experiments to production workflows across entire organizations. The constraint is no longer the models themselves, but the infrastructure to orchestrate them, the workforce skills to deploy them, and the platforms to make them portable across tools. This week, watch for more announcements about forward-deployed engineers, enterprise partnerships, and security-focused AI tools. The real competition is not model benchmarks, but who can help businesses actually deploy AI at scale.