MCP Is Becoming the USB-C of AI Agents
EU's AI Omnibus Takes Effect, Trades Some Friction for Speed.
MCP Is Becoming the USB-C of AI Agents
The Model Context Protocol, Anthropic's open standard for connecting models to external tools, databases, and apps, has moved from interesting proposal to default infrastructure [1][2]. Originally introduced in late 2024, MCP now has real traction: Cursor, Claude Code, and a growing list of enterprise stacks support it, and Microsoft has published its own guidance for building MCP-based agents on Azure [3].
The appeal is obvious once you've built anything agentic: instead of writing bespoke integrations for every calendar, repo, or database an AI needs to touch, you get one standard, two-way connection protocol. Local servers for dev work, remote servers for scale. It's plumbing, but it's the plumbing that decides whether "AI agent" is a demo or a production system.
X threads this week focused on commerce and app-interaction use cases — agents that don't just answer questions but execute multi-step tasks across tools. That's the shift worth watching: MCP adoption is a leading indicator of how seriously a company is building for orchestration rather than one-off chatbot features.
EU's AI Omnibus Takes Effect, Trades Some Friction for Speed
On July 27, 2026, the EU's AI Omnibus officially entered into force, softening parts of the AI Act's compliance load while keeping core protections intact [1][2]. Startups get extended timelines and simplified admin requirements; child protection rules, bias detection standards, and bans on tools like nudification apps stay firmly in place. It follows provisional agreements reached in May and June, and Brussels has an "innovation month" of events planned to sell the story that Europe is serious about competitiveness, not just regulation.
For Nordic AI builders, this is the practical news of the week. The AI Act was never the problem — the uncertainty and paperwork around it was. Loosening the administrative burden while keeping the guardrails is a rare case of a regulator listening to the "we can't out-innovate the US if compliance eats our runway" argument without gutting the protections that build public trust in AI systems. X engagement reflects relief more than celebration: companies aren't thrilled about AI Act 2.0, they're relieved it's more workable than feared.
What This Means For Your Business
Three stories, one thread: the value is moving from writing code to directing systems that write, connect, and act on their own. Anthropic teaching prompting at scale, MCP becoming the standard wiring for agents, and Europe simplifying compliance so companies can actually ship — all of it points toward an environment where the constraint isn't "can we build this" but "do we know what we want it to do, and can we trust it to do it right."
That's judgment, not code. The teams winning right now aren't the ones with the most engineers — they're the ones who've figured out how to specify outcomes clearly, wire tools together via standards like MCP, and verify agent behavior before it touches production. If your organization is still hiring for "can write Python" instead of "can architect a reliable agent system and know when to intervene," you're optimizing for the wrong decade.
The regulatory softening in the EU removes an excuse. Compliance friction was a legitimate reason to move slowly before; it's a weaker one now. Combine that with free, high-quality training material from the model providers themselves, and there's no structural reason left for a Nordic company to sit on the sidelines of agentic AI — only organizational ones.
Key takeaway: The tools to orchestrate AI are now free, standardized, and increasingly de-regulated in Europe — the only remaining bottleneck is whether your team has the judgment to use them well.
Sources
- https://www.anthropic.com/learn
- https://www.anthropic.com/learn/claude-for-you
- https://github.com/anthropics/courses
- https://modelcontextprotocol.io/docs/getting-started/intro
- https://www.anthropic.com/news/model-context-protocol
- https://learn.microsoft.com/en-us/azure/developer/ai/intro-agents-mcp
- https://digital-strategy.ec.europa.eu/en/news/ai-omnibus-enters-force
- https://digital-strategy.ec.europa.eu/en/library/digital-omnibus-ai-regulation-proposal
Stay ahead of AI
No spam. Unsubscribe anytime.
Want to go deeper?
Reading the news is one thing. Exploring the frontier is another. See what we're building.