Summary

Federal Chancellor Viktor Rossi met with researchers from 6 universities, 3 universities of applied sciences, and the Zurich Forensic Institute to discuss measures to strengthen the integrity of signature collections for popular initiatives and referendums. The scientists developed concrete proposals, including the use of artificial intelligence to detect duplicate signatures and forged handwriting. Several measures have already been implemented, while the Federal Chancellery will inform about further steps.

People

Topics

  • Signature Collections & Integrity
  • Artificial Intelligence in Administration
  • Popular Initiatives and Referendums
  • Fraud Prevention

Clarus Lead

The Federal Chancellery is working systematically with the scientific community to prevent manipulation of signature collections. An exchange that has been ongoing since autumn 2024 with nine universities and research institutes has resulted in concrete improvement proposals. Particularly relevant: Artificial Intelligence is to automatically detect duplicate signatures and forged handwriting patterns in the future – while the final decision remains with humans. For decision-makers, this means a modernization of control mechanisms in direct democratic processes.

Detailed Summary

The exchange between the Federal Chancellery and researchers from a total of nine educational institutions as well as the Zurich Forensic Institute has produced a broad range of measures. The focus was on two strategies: on the one hand, organizational and preventive measures such as training of counting teams, raising awareness among initiative committees, and improved guidelines. On the other hand, the technological dimension was examined – in particular the question of whether artificial intelligence can reliably uncover fraud attempts.

The research group successfully tested whether AI systems can detect duplicate signatures and recurring handwriting patterns. This tool is intended to supplement the Federal Chancellery's manual controls, not replace them. Several measures have already been implemented, such as training counting teams. At the same time, the Federal Chancellery has launched a voluntary code of conduct, introduced stricter controls, and established a nationwide monitoring system through which municipalities and cantons can report irregularities. Since 2022, five criminal complaints have been filed.

Key Statements

  • Science-Practice Cooperation: Nine universities and institutes jointly developed concrete improvement measures with the Federal Chancellery.
  • AI as a Control Instrument: Artificial intelligence can automatically detect duplicate signatures and handwriting forgeries, but does not itself decide on validity.
  • Multi-Level Measure Package: In addition to technical solutions, training, codes, guidelines, and a monitoring system have been implemented.
  • Enforcement: Since 2022, five criminal complaints have been filed; further suspected cases are being pursued consistently.

Critical Questions

  1. Data Quality & Validation: How reliable are the AI tests to date? Have detection rates been verified in representative samples, and what is the rate of false alarms?

  2. Conflicts of Interest & Independence: To what extent were the researchers involved independent of political expectations? Were success criteria for AI evaluation predetermined?

  3. Causality & Alternatives: Can it be demonstrated that the measures taken so far (code of conduct, training, monitoring system) have actually led to fewer fraud cases? Which alternative technologies were considered?

  4. Feasibility & Risks: How is it ensured that AI systems are applied uniformly across all cantons and municipalities? What data protection and security risks arise from the digitization of signature data?

  5. Transparency & Verifiability: Are the criteria by which the AI classifies a signature as suspicious publicly documented and can they be understood by initiative committees?

  6. Scalability: How many signature collections per year can be checked with the new systems, and where are the capacity limits?

Source Directory

Primary Source: Integrity of Signature Collections: Federal Chancellor Meets Researchers for Exchange – Press Release Federal Chancellery, February 12, 2026

Verification Status: ✓ February 12, 2026

This text was created with the support of an AI model. Editorial Responsibility: clarus.news | Fact-Check: February 12, 2026