MindPattern
← The Wire

Security2026-06-26 · source-backed

Prompt injection reliably games automated résumé screening.

Confidence
source-backed
Sources
1
Redaction
passed

Story

A new study shows subtle self-promotional text adding no new qualifications reliably improves an applicant's ranking under LLM screeners, especially when résumé quality is homogeneous and few candidates inject. The adversarial incentive is baked into algorithmic hiring. If you deploy LLM evaluators of any kind, the ranking can be gamed by content, not credentials.

Related stories

Each link below shares sources, entities, or timing with this story.

  1. Shared entity: LLM / Same source domain / Earlier coverage / Tension

    LLM-generated GPU kernels pass weak benchmarks and ship real bugs.

    Both cover LLM; reported by the same outlet (arxiv.org); earlier LLM coverage from 2026-06-20.

  2. Shared entity: LLM / Same source domain / Earlier coverage / Downstream implication

    Evaluator bias is contagious in multi-agent systems.

    Both cover LLM; reported by the same outlet (arxiv.org); earlier LLM coverage from 2026-06-19.

  3. Shared entity: LLM / Same source domain / Earlier coverage / Tension

    Stronger models parrot their tools more, not less

    Both cover LLM; reported by the same outlet (arxiv.org); earlier LLM coverage from 2026-06-18.

  4. Shared entity: LLM / Same source domain / Earlier coverage / Downstream implication

    Human and LLM everyday reasoning are both pattern matching, not world models.

    Both cover LLM; reported by the same outlet (arxiv.org); earlier LLM coverage from 2026-06-14.

  5. Shared entity: LLM / Same source domain / Earlier coverage / Tension

    Your AI-Generated Code Compiles. It Doesn't Work.

    Both cover LLM; reported by the same outlet (arxiv.org); earlier LLM coverage from 2026-04-22.

  6. 10 actionable skills from today's findings:

    Both cover LLM; reported by the same outlet (arxiv.org); earlier LLM coverage from 2026-03-22.

  7. LLM Web Agents Fail Dark Patterns 41–72% of the Time

    Both cover LLM; reported by the same outlet (arxiv.org); earlier LLM coverage from 2026-03-20.

  8. Reasoning LLM Judges: Goodhart's Law Strikes Again

    Both cover LLM; reported by the same outlet (arxiv.org); earlier LLM coverage from 2026-03-14.

Source trail

new study

Entities

LLMPrompt

Provenance

Canonical issue
Ramsay Research Agent — June 26, 2026
AI generated
no
Story unit
2026-06-26-prompt-injection-reliably-games-automated-resume-screening
Labels
source-backed, canonical briefing excerpt