Summary

The Swiss Cyber Command has released Version 1.0.0 of the open-source software "Loom". The document analysis platform automatically indexes large amounts of data from various formats (Office, PDF, emails, archives, images) and enables AI-powered search as well as summaries. The software runs completely locally and offline, ideal for organizations with confidential or classified data. Loom was developed significantly by members of the Cyber training courses. The software is now registered in the Federal Chancellery's open-source catalogue and is available under MIT license via GitLab.

People

  • Marco Remund (Communications, Cyber Command)

Topics

  • Open-source software
  • Artificial intelligence
  • Digital sovereignty
  • Federal administration

Clarus Lead

The release of Loom signals a strategic opening of the Swiss Armed Forces: a military cyber command makes developed technology available to the public and thereby underscores the "Public Money, Public Code" principle. This aligns with the Federal Law on Electronic Government Services (EMBAG) and positions Switzerland as an actor for digital transparency and sovereignty. This creates a multiplier effect for the federal administration: other authorities and private organizations can adapt and further develop Loom without paying license fees.

Detailed Summary

Functionality and applications: Loom automatically indexes heterogeneous data sources and extracts structured content and metadata. An AI-powered chatbot answers questions based on indexed documents, creates summaries, and translates texts. Scanned documents are made searchable through optical character recognition (OCR). The offline operating model makes the software particularly attractive for authorities and organizations with data protection or security requirements.

Roadmap and technical development: Further development focuses on advanced AI features: image recognition capabilities should automatically generate descriptions for uploaded images, making visual content searchable in text form. Another goal is automatic document categorization. In parallel, integration of external storage systems is being developed so that organizations can index existing data directly from their infrastructures. The processing pipeline is continuously optimized – focus on memory efficiency and support for additional file formats.

Militia as a success factor: A large portion of contributors come from refresher courses (WKs) and cyber training course participants. Their civilian and military expertise enabled efficient problem-solving and innovative implementation. This model of collaboration is highlighted as a success factor for technological developments within the military.

Key Findings

  • Loom 1.0.0 is a locally operated document analysis platform with AI chatbot, OCR, and translation function.
  • The software runs completely offline and thus protects classified and confidential data.
  • Publication under MIT license follows the "Public Money, Public Code" principle and EMBAG.
  • Militia members were instrumental in development.
  • Software is now registered in the Federal Chancellery's open-source catalogue.

Critical Questions

  1. Evidence/Data Quality: How was the accuracy of the OCR function tested across different document types and languages? What error rates exist?

  2. Conflicts of Interest/Independence: To what extent do military requirements influence feature prioritization compared to civilian use cases?

  3. Causality/Alternatives: What other open-source solutions exist in the document management field, and how does Loom functionally differ from them?

  4. Feasibility/Risks: What security audits were conducted before release? How are critical security vulnerabilities handled?

  5. Storage and Performance Limits: At what data volume does performance decline significantly? Are there benchmarks?

  6. Adoption: Which federal authorities are already planning to use Loom?

Source Directory

Primary Source: VTG – Cyber Command publishes version 1.0.0 of open-source software Loom – https://www.vtg.admin.ch/de/newsvtg-kommando-cyber-veroeffentlicht-version-100-der-open-source-software-loom

Supplementary Resources:

  1. Federal Chancellery Open-Source Catalogue: https://opensource.admin.ch
  2. GitLab Repository (source code under MIT license)

Verification Status: ✓ 2025

This text was created with the support of an AI model. Editorial Responsibility: clarus.news | Fact-checking: 2025