[ Resources ]

Naive vs Paperclip: Naive is Not Paperclip

Our approach to attribution, open source, and the infrastructure we're building.

April 8, 2026·Dennis Zax, CTO

TL;DR

Attribution exists. Paperclip is credited by name in our Terms of Service §7.4 — with repo link, MIT license, and copyright notice.
Nothing was hidden. Residual strings in a minified bundle are standard open-source artifacts, not evidence of concealment.
Not a wrapper. Naive's core IP is the identity & primitives layer — LLCs, bank accounts, insurance, comms, tool access — plus our own runtime, self-improvement layer, and template ecosystem. None of this exists in Paperclip.
Committed to open source. We are actively building and releasing open-source tooling ourselves.

Attribution fully exists — and always has

The claim that we provide “zero attribution” is false. Section 7.4 of our Terms of Service explicitly states:

“The Service incorporates open source software components, including Paperclip, licensed under the MIT License. Copyright © 2025 Paperclip AI. Full license terms are available in the project repository and on our Open Source Licensespage. Nothing in these Terms restricts your rights under applicable open source licenses.”

This section has been in our Terms since we launched. We link directly to Paperclip's GitHub repository, name the MIT license, and reproduce the copyright line. That is the attribution the MIT license requires. Paperclip is the most visible, but it is one of many open-source projects we build on — and we attribute each of them in the same way.

No intent to hide anything

We have never attempted to conceal Paperclip's role in our stack. The leftover “Paperclip” references in our production bundle are exactly what you'd expect when building on top of a large open-source project — string literals, CSS class names, and config defaults that flow through from upstream. That's how open-source software works: you build on top of it, and not every internal string gets rebranded.

Framing inherited code artifacts as “smoking guns” misrepresents how software is built. Every major SaaS product ships with hundreds of upstream references in their bundles. It is not evidence of deception — it is evidence of usage.

What Naive actually builds

Paperclip is an agent orchestration and runtime framework — and a good one. We chose to build on it because it's MIT-licensed, well-engineered, and aligned with the direction of agentic infrastructure. We have enormous respect for Dotta and the Paperclip community.

But Naive is not “a billing layer wrapped around Paperclip.” Our core product — the part we've built and continue to build — is the identity and primitives layer that makes agents actually operational in the real world:

At its core, Naive is the identity and primitives layer for agents. We are building exclusive partnerships that enable agents to:

  • Provision LLCs: Agents can form and manage real legal entities.
  • Create and manage bank accounts: Real bank accounts with payment cards and per-agent budget enforcement.
  • Get insurance: Agents can obtain coverage for the businesses they operate.
  • Manage email and phone numbers: Dedicated communication channels — agents send, receive, and respond to real messages.
  • Sign up and pay for tools/services: Authenticated browser sessions with 2FA support, so agents can operate real SaaS products and pay for them autonomously.

This primitives layer is agent runtime and orchestration independent. It works with Paperclip today — and it would work with any orchestration framework tomorrow.

Beyond primitives, we also build our own agent runtime and self-improvement layer — the feedback loops that let agents learn from their actions, evaluate outcomes, and get better over time. And we maintain a growing ecosystem of templatesthat let anyone deploy pre-configured companies and workflows in minutes. These are all things we've built ourselves, on top of the open-source foundation.

We believe in open source

We are not adversaries of open source — we are practitioners. We are actively building and open-sourcing primitives and tooling for the agentic ecosystem. We believe the agent infrastructure layer should be composable and open, and we intend to contribute meaningfully to that ecosystem, both upstream and through our own projects.

Using MIT-licensed software commercially is not a violation of open-source norms — it is the explicit intent of the MIT license. We comply with its terms, we attribute the project, and we are grateful for the foundation it provides.

Our Team

We're a small team building fast. We're not going to get everything right, and we welcome good-faith criticism. But the narrative that we “stripped attribution” and are “presenting someone else's work as our own” is factually incorrect. Attribution is in our Terms. Our product goes well beyond any single upstream dependency. And our commitment to open source is genuine and ongoing.

Questions? Reach us at dennis@usenaive.ai