Paper sets an accountability test for autonomous cyber weapons
Original title: Autonomous Cyber Weapons as AI-Enabled Digital Harms: A Platform Accountability Framework
On 11 documented operations, none cleared the authors' civilian-target human-control threshold; weights are illustrative.
The paper argues nation-state autonomous cyber weapons aimed at civilian infrastructure fall outside platform regimes such as the EU DSA, the AI Act and the UK Online Safety Act. It offers a harm taxonomy of six weapon modes and five harm categories, a Meaningful Human Control function with a stated international-humanitarian-law threshold, and platform-accountability language grounded in ARSIWA state responsibility rather than the Rome Statute. Applied to eleven documented operations and one hypothetical defensive system, the authors say every documented case sits in an accountability vacuum and no studied system clears the bar against wholly civilian targets.
Unconfirmed: Threshold parameters are illustrative, one system is hypothetical, and case outcomes are the authors' application.
- Three further incidents were excluded on autonomy or attribution grounds.
- The control threshold is described as target-dependent.
- Weights and threshold parameters are illustrative, not fitted.
- The design uses state responsibility under ARSIWA, not individual criminal liability.
From the original article
1 min read at arxiv.orgGovernments govern AI-generated online harms from commercial platforms (EU DSA, AI Act, UK Online Safety Act), yet the most severe class remains ungoverned: autonomous cyber weapons (ACWs) deployed by nation-states against civilian infrastructure.
By Mansi Ganga Nayak · arxiv.org
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.