Mozilla launches Thunderbolt AI client targeting enterprises seeking data sovereignty

Mozilla enters enterprise AI market with sovereignty focus

Mozilla has entered the enterprise AI market with Thunderbolt, a new open-source AI client designed for organizations seeking control over their AI infrastructure. Unlike competitors offering cloud-based services, Thunderbolt positions itself as a privacy-focused, self-hosted alternative to enterprise AI platforms like Microsoft Copilot, ChatGPT Enterprise, and Claude Enterprise. The product is being developed by MZLA Technologies, the Mozilla Foundation subsidiary formed in 2020 to manage the Thunderbird email client1

1

.

Rather than releasing a standalone AI model, Mozilla built Thunderbolt as a front-end client that operates on top of self-hosted AI infrastructure. MZLA CEO Ryan Sipes framed the launch as addressing fundamental questions of sovereignty and control. "Do you really want to build your AI workflows on top of a proprietary service from OpenAI or Anthropic ... not to mention having all your internal company data flowing through their systems?" Sipes told The Register3

3

.

Built on Haystack framework with extensive compatibility

Thunderbolt is built on top of Haystack, an existing open-source AI framework from German company deepset that enables users to build custom, modular AI pipelines from user-chosen components1

1

. Acting as what Mozilla calls a "sovereign AI client," Thunderbolt can plug into any Agent Client Protocol (ACP)-compatible agent or OpenAI-compatible API, including Claude, Codex, OpenClaw, DeepSeek, and OpenCode1

1

.

The AI execution environment integrates with locally stored enterprise data through open protocols and uses an offline SQLite database as a local "source of truth" for the model to reference1

1

. This architecture allows organizations to maintain complete control over their entire AI service stack, addressing concerns about data leaking to outside providers. For free local inference, Mozilla recommends using Ollama or llama.cpp, though other OpenAI-compatible model providers can be configured by setting an API key5

5

.

Data privacy and security at the core

Data privacy and security features distinguish Thunderbolt in the crowded enterprise AI landscape. Mozilla says the platform offers "optional end-to-end encryption, and device-level access controls" for additional security1

1

. Sipes emphasized that the system can be configured to run in environments as small as a single machine if sensitive data needs protection3

3

.

The emphasis on data sovereignty reflects growing enterprise concerns about vendor lock-in and data control. "When you rely on these big proprietary providers, you're just renting a critical part of your organization's operations," Sipes explained. "Whereas if you deploy Thunderbolt and use open source agents like what can be created via deepset's Haystack platform - you own your AI stack, end-to-end"3

3

.

Cross-platform availability and enterprise readiness

Thunderbolt supports familiar AI interfaces including chat, search, research, automation, and cross-device workflows1

1

. Native applications for Windows, macOS, Linux, iOS, and Android are available for direct download or can be built from React source code via a GitHub repository2

2

. Organizations can automate workflows and recurring tasks such as generating daily briefings, monitoring topics, compiling reports, or triggering actions based on events and schedules5

5

.

Mozilla is already encouraging potential enterprise clients to coordinate paid licensing and on-site deployments, though the GitHub page warns that Thunderbolt is "under active development, currently undergoing a security audit, and preparing for enterprise production readiness"1

1

. While designed to be completely offline-first in the future, the current version still relies on authentication and search functions, though search can be disabled and local testing is possible by running a custom backend with Docker5

5

.

Mozilla's broader AI strategy and market positioning

Thunderbolt builds on Mozilla's existing AI efforts through Mozilla.ai, which backs open source tooling for external AI models and agents. In late 2025, Mozilla announced its aim to "do for AI what we did for the web" by offering AI services with more agency, diversity, and user choice through a "decentralized open source AI ecosystem that matches the capabilities of Big AI"1

1

.

Sipes drew parallels to Firefox's challenge against Internet Explorer's 95% market share in the early 2000s. "We, collectively, beyond just Mozilla, have to create alternatives to Copilot and ChatGPT so that the future of AI isn't just us renting it from a few gigantic companies," he explained3

3

. The MZLA CEO positioned Mozilla as "part of the rebel alliance" attempting to build alternatives in what's becoming a market dominated by massive companies. The code is released under the Mozilla Public License (MPL)-2.0 license5

5

, reinforcing Mozilla's commitment to open-source principles and giving organizations the freedom to deploy Thunderbolt within their own infrastructure without ongoing vendor relationships.