Entity · organization

University of Technology Sydney

organizationactiveprovisionaluniversity-of-technology-sydney-fe8a2ac5·1 events·first seen 5d ago

Aliases: University of Technology Sydney

Co-occurring entities

ELOQUENT 2026 Voight-Kampff UTS at ELOQUENT 2026 Voight-Kampff: structural shifts in AI writing bypass state-of-the-art detectors

More like this (12)

Hong Kong University of Science and Technology University of Toronto Nanyang Technological University New York University University of Saskatchewan Technical University of Munich University of Texas Austin University of Bristol Curtin University Australian National University Technology Innovation Institute University of Cambridge

Recent events (1)

6arXiv · cs.CL·5d ago·source ↗

Structural out-of-distribution attacks bypass adversarially fine-tuned AI text detectors at ELOQUENT 2026

Researchers from UTS demonstrate that adversarial fine-tuning of AI text detectors, while closing 2025-era evasion methods, fails against a newly identified class of out-of-distribution attacks. Two novel attack families — cross-decade register shifts and modernist stream-of-consciousness form — achieve up to ~50x higher fool rates than prior methods while preserving text naturalness. The work reveals a fundamental asymmetry: pushing generated text out of a detector's training distribution reliably defeats detection, while pulling it toward human training data does not. These findings swept the top 5 positions on the ELOQUENT 2026 Voight-Kampff leaderboard and suggest persistent structural vulnerabilities in current detector families.

Evaluation and Benchmarking AI Safety Research ELOQUENT 2026 Voight-Kampff UTS at ELOQUENT 2026 Voight-Kampff: structural shifts in AI writing bypass state-of-the-art detectors University of Technology Sydney