aiqbench analysis
This research introduces DualEvasion, a new benchmark that evaluates multimodal AI models on detecting corporate evasion through both text and vocal confidence cues in earnings calls. The study reveals significant limitations in current state-of-the-art systems, which struggle to interpret acoustic signals relative to individual speaker baselines rather than in isolation. This work highlights critical gaps between automated detection capabilities and human performance in analyzing the multidimensional nature of spoken communication.
A Shaky Voice Is Not Always a Dodge: Benchmarking Textual and Vocal Evasion Detection in Earnings Calls
arXiv:2608.28040v1 Announce Type: new Abstract: Existing approaches to evasion detection in earnings calls focus on textual transcripts, treating evasion as a single-dimensional phenomenon. We argue that evasion in spoken communication is inherently multidimensional: beyond what executives say, how they say it carries independent and complementary information. To study these dimensions jointly, we introduce DualEvasion, a benchmark for evasion detection across text and audio in earnings call Q&A. The benchmark contains 505 annotated question-answer pairs from 60 earnings calls, each with two independent labels: textual evasion (direct vs. evasive) and vocal cues operationalized as speaker confidence (confident vs. unconfident). Our experiments show that state-of-the-art multimodal models struggle to detect vocal confidence, particularly on unconfident responses. Our analysis suggests these models interpret acoustic cues in isolation rather than relative to each speaker's baseline. Providing speaker-level references yields modest improvements, but a substantial gap with human performance remains.
Abstract:Existing approaches to evasion detection in earnings calls focus on textual transcripts, treating evasion as a single-dimensional phenomenon. We argue that evasion in spoken communication is inherently multidimensional: beyond what executives say, how they say it carries independent and complementary information. To study these dimensions jointly, we introduce DualEvasion, a benchmark for evasion detection across text and audio in earnings call Q&A. The benchmark contains 505 annotated question-answer pairs from 60 earnings calls, each with two independent labels: textual evasion (direct vs. evasive) and vocal cues operationalized as speaker confidence (confident vs. unconfident). Our experiments show that state-of-the-art multimodal models struggle to detect vocal confidence, particularly on unconfident responses. Our analysis suggests these models interpret acoustic cues in isolation rather than relative to each speaker's baseline. Providing speaker-level references yields modest improvements, but a substantial gap with human performance remains.
Computation and Language (cs.CL); Sound (cs.SD)
arXiv:2608.28040 [cs.CL]
(or arXiv:2608.28040v1 [cs.CL] for this version)
https://doi.org/10.48550/arXiv.2608.28040 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 237 words
Related coverage
[AINews] The Future of Latent Space
Latent Space · Sep 25, 2026
Microsoft's new AI model beats Mythos on security benchmark
ZDNet · Jul 28, 2026
MUUNRiver-Bench: Diagnosing Relation-Dependent Music Retrieval with Multimodal Instructions
arXiv · Sep 16, 2026
Links #5: 2026/07
LessWrong · Jul 31, 2026