AWS just made a quiet move that says a lot about where enterprise AI is heading. Vibe-coding startup Superblocks announced a multiyear joint marketing agreement with Amazon Web Services that lets its tool run inside the private clouds of AWS customers, according to TechCrunch AI. In plain terms: a company on AWS can now offer vibe coding to its business users, and the apps those users build never send data outside the company’s own cloud.
That last part is the whole point.
What actually changed
Until now, vibe coding tools like Lovable and Replit have lived mostly outside the enterprise security perimeter. A business user spins up an app, and it might create an external Supabase database or route data to a model provider you don’t control. IT has no visibility. Those are the “rogue apps” every CIO worries about.
Superblocks flips that. As TechCrunch AI reports, apps built through the AWS integration will:
- Spin up Amazon Aurora databases inside the company’s private cloud, not external ones
- Integrate with Amazon Bedrock, AWS’s AI gateway and inference platform
- Fall automatically under IT’s management, security, auditing, and encryption
“We’re going to bring it to your data inside your private cloud,” Superblocks co-founder and CEO Brad Menezes told TechCrunch. “The big thing about that is data never leaves. It’s their AWS account and basically secure with all of the auditing, all of the encryption, all of the network controls.”
AWS will also help sell Superblocks to enterprises, the same way it pushes other Marketplace partners.
Why this matters beyond one startup
Superblocks is small. It has 50 employees and has raised $60 million total, backed by Spark Capital, Kleiner Perkins, Meritech, and Greenoaks. On its own, this deal is a nice boost for an early-stage company.
What stands out here is the pattern behind it.
The hyperscalers are pushing enterprises to separate the AI model from everything wrapped around it. The orchestration, the security tooling, the agentic app layer, the “harness” that turns a model into a working business app. Cloud providers want you buying all of that scaffolding from them, not from the frontier labs like OpenAI or Anthropic.
Microsoft CEO Satya Nadella has been making the same argument in public for weeks. He’s told enterprise customers to run multiple models to cut costs and dodge lock-in, and he’s warned that the AI labs may not be trustworthy enough to handle agent orchestration, since they could study your business and later compete with it.
The multi-model shift is real
Enterprises don’t seem to need much convincing. Menezes says the mood flipped fast.
“That is flipped because 60 days ago they were like, I want a specific model. It’s called Anthropic,” he told TechCrunch.
Now the demand is model choice across coding, customer service, HR, and sales automation.
The data backs him up. Open models accounted for 29% of all traffic routed through Vercel’s AI gateway last month, a tool enterprises use to juggle multiple models. Chinese open-weight options are climbing, and U.S. open source is picking up too.
“Any enterprise that is betting on a single model provider, that executive will be fired.”
Where AWS still has a gap
Here’s the interesting wrinkle. AWS doesn’t yet have its own vibe-coding agent for business users. It has Kiro, a coding agent aimed at developers, and Quick, an AI assistant that’s closer to Microsoft Copilot than to Lovable or Replit.
So AWS is filling a real hole in its lineup through a partner rather than building it in-house. That’s why the deal reads as a second wave. First the cloud giants brought AI coding agents to enterprise developers. Now they’re bringing vibe coding for business users into private, secure clouds.
“It’s an emerging category with real momentum, and exactly the kind of innovation we support,” an AWS spokesperson told TechCrunch.
What to watch next
Expect the other hyperscalers to answer. If AWS is happy to route around its own product gap with a partnership, Microsoft and Google have every reason to lock down the same category before frontier labs sell app-layer harnesses directly to CIOs. The battle isn’t really about who has the best model anymore. It’s about who owns the secure plumbing around it. You can find the full report at TechCrunch AI.