Dust: AI Agent Orchestration and Secure Enterprise Data Integration

Dust is a Paris-based deep-tech startup building a secure AI “operating system” for enterprises, enabling automated agents to access and act on internal corporate data and workflows [1] [2] . Founded in 2023 by former Stripe and OpenAI engineers (Gabriel Hubert and Stanislas Polu [3] [4] ), Dust’s platform lets companies spin up specialized AI assistants integrated with tools like Slack, Salesforce and GitHub, processing knowledge bases without manual effort [1] [5] . This vision – moving beyond standalone chatbots to collaborative, task-oriented AI agents – positions Dust at the forefront of the enterprise AI revolution. In the context of Europe’s drive for digital sovereignty and defense readiness, Dust’s technology exemplifies how advanced AI innovation is taking root in Europe.

NATO’s 2022 Strategic Concept emphasizes strengthening deterrence and investing in cutting-edge technology [6] [7] , and the EU’s Defence Fund explicitly identifies AI and data integration as critical capabilities [8] [9] . Dust’s secure, data-centric AI platform taps directly into these priorities, even as it remains outside traditional defense supply chains. The analysis below examines Dust’s technical strengths, strategic positioning for European autonomy, and potential gaps relative to NATO/EU defense needs. Dust is a Paris-headquartered deep-tech startup specializing in AI-powered workflow automation.

Its core technology aligns with European EDT priorities in Artificial Intelligence & Machine Learning (military AI/decision support) and Cybersecurity (secure data access). Unlike consumer AI tools, Dust’s agents are connected to corporate data (documents, code, CRM) and operate autonomously to perform tasks [1] [10] . This capability corresponds to NATO/EDT goals for digital enablers: AI that enhances information processing, situational awareness and decision-making [8] [1] . While Dust is not a traditional defense contractor, it contributes to European sovereignty by growing a homegrown AI platform.

The company’s technology does not depend on Chinese AI providers (supporting EU autonomy [9] ) but rather relies on Western models (OpenAI, Anthropic) integrated locally.