Insights on AI and Recruitment
Three studies in 2025. One structural defect. A different architecture is now table stakes.
Three studies in 2025 documented the same structural defect: LLMs systematically prefer text they wrote themselves. If your hiring stack scores resumes with an LLM and your candidates write resumes with an LLM, you are not running a screening process.
A new paper from one of the world's most respected mathematicians introduces a framework from cybersecurity that explains why so many AI hiring tools are failing in court, in audits, and in front of regulators.
Klowden and Tao apply a cybersecurity frame to AI. Red team verifies. Blue team generates structural decisions. Most AI hiring tools are blue team. The settlements, regulations, and class certifications are downstream of that architecture choice.
The 9-to-5 phone screen systematically excludes a third of the modern workforce. The evidence is in who calls back at 11pm.
How structured AI pre-screening with proper inbound channels reaches the talent pool the synchronous workday never sees. Vena Medical pilot: 98% inbound completion rate vs. 37% outbound, a 2.6× lift, and what it means for the hiring funnel as a category.
New research introduces 'cognitive surrender' — when people stop thinking and just accept AI outputs. The findings have direct implications for how AI should be used in hiring decisions.
New research introduces 'cognitive surrender' — when people stop thinking and just accept AI outputs. The findings have direct implications for how AI should be used in hiring decisions.
Harvard and Burning Glass found that dropping degree requirements changed fewer than 1 in 700 hires. Here's what's really happening and what it takes to close the gap.
Harvard and Burning Glass found that dropping degree requirements changed fewer than 1 in 700 hires. Here's what's really happening and what it takes to close the gap.
Checkr's survey of 3,000 managers found 60% have caught candidates lying about qualifications. 31% have interviewed people using fake identities. Here's how cross-verification catches what traditional checks miss.
Checkr's 2025 survey of 3,000 managers revealed 60% have caught candidates lying about qualifications. 31% have interviewed candidates using fake identities. Traditional verification methods are failing — here's how cross-verification catches what single-source checks miss.