arXiv:2606.15815v1 Announce Type: new Abstract: The deployment of NLP systems has raised concerns about harms they might produce, including representational harms. Recent literature has begun to conceptualize and measure one such harm, the harm of erasure. Nevertheless, the field lacks a clear and cohesive conceptual foundation for identifying and measuring erasure. Existing conceptualizations of erasure are often broad -- making it difficult to identify what is needed to establish and measure erasure -- or else specific to particular settings -- facilitating measurement for those settings but
The increased deployment and societal impact of NLP systems necessitate a more robust understanding and mitigation of potential harms, such as erasure.
A clear conceptual framework for identifying and measuring 'erasure harms' allows for more responsible and ethical AI development, mitigating future societal and regulatory friction.
The explicit definition and measurement of 'erasure harms' will lead to new metrics, ethical guidelines, and potentially regulatory requirements for NLP system developers.
- · AI ethicists
- · NLP researchers focused on fairness
- · Users vulnerable to representational harms
- · Responsible AI developers
- · Developers ignoring ethical AI principles
- · Companies facing reputational damage from biased systems
- · Rapid, unregulated deployment of NLP systems
The NLP community will adopt standardized definitions and measurement techniques for erasure harms.
New tools and frameworks will emerge to detect, quantify, and mitigate erasure in commercially deployed NLP systems.
Future regulations and compliance standards for AI systems will likely incorporate specific requirements related to the prevention of representational harms like erasure.
This signal links to a primary source. Continuum Brief monitors and indexes it as part of the live intelligence stream — we do not republish source content.
Read at arXiv cs.CL