Field notes from cleaning hundreds of thousands of resumes for recruitment agencies. Practical, no fluff.
New research reveals how adversarial CV content can exploit recruitment AI tools through prompt injection—and why 31 of 36 real-world applications failed to defend against it.
Read articleNew research shows 0.1% poisoned training data can hijack AI decisions. Learn how data-poisoning attacks threaten recruitment AI and what cv-cleaner does to protect you.
Read articleResearch reveals all LLM-based screening tools are vulnerable to prompt injection attacks—and current defenses fail. Here's what recruiters need to know.
Read articleAttackers can remotely hijack AI recruiting systems by embedding malicious prompts in CVs—bypassing filters, stealing data, and manipulating hiring outcomes without direct access.
Read articleMedical AI research reveals how hidden instructions can manipulate vision-language models—exposing the same vulnerability that threatens CV screening tools used by recruiters.
Read articleNew research reveals how attackers manipulate AI screening tools through prompt injection—and how recruitment teams can detect these hidden threats in real-time.
Read articleNew research shows AI systems give harmful responses 15-30% of the time when fed malicious prompts—a finding that directly threatens fair, compliant recruitment.
Read articleNew research reveals AI agents are vulnerable to hidden instructions in external content—including CVs. Learn how prompt injection threatens recruitment AI and what it means for hiring security.
Read articleService-oriented architecture shows how integration platforms handle complexity, security, and data transformation—lessons cv-cleaner applies to inbound CV workflows.
Read articleHow mechanical, thinking, and feeling AI transform recruitment workflows—from CV processing to candidate engagement and fair hiring decisions.
Read articleHow treating fairness as a robustness problem helps recruitment AI deliver consistent, unbiased candidate evaluations—even when sensitive attributes change.
Read articleInvisible white-on-white keyword stuffing has quietly replaced classic resume lies. Why it works, why it matters, and how to defend your screening pipeline.
Read articleLawful basis, retention windows, processor agreements, candidate rights. A practical GDPR primer for EU agencies handling CVs in 2026.
Read articleWorkday, Greenhouse, Lever, Bullhorn, Taleo - what each ATS actually accepts in 2026. A reference for recruitment agencies submitting candidate CVs.
Read articleBlind hiring reduces bias only when anonymization is consistent. A 12-point checklist for recruitment agencies running structured anonymized CV pipelines.
Read articleCandidates are embedding hidden instructions in resumes to manipulate AI-based screening. Here's how recruitment agencies can detect and neutralize prompt injection in CVs.
Read article