AI screening still needs humans
A March 25 podcast highlighted a case where Sapia's AI screening mischaracterized a 16‑year‑old job candidate, reigniting the argument that AI is a first‑pass filter but can't replace human assessment for nuanced fit — a practical warning for board recruiting firms using automated tools. (omny.fm) (boardsi.com) (hrkatha.com)
The Matt Heath & Tyler Adams episode dated March 25, 2026 ran about 116 minutes and lists a segment on CVs and AI in recruitment in its published episode notes. (iHeart). Sapia publicly launched its SAIGE interview-grading product in a March 13, 2024 press release and has since marketed real‑time detection of AI‑generated candidate answers as a platform capability. (BusinessWire; SiliconCanals). Sapia has cited a proprietary corpus in its product announcements — a 2023 release referenced “over one billion words” from millions of responses, while later partner and trade pieces describe the dataset growing to roughly 2 billion words and tens of millions of interview pairs. (BusinessWire; TechRSeries). Company materials and review sites list large enterprise customers — Sapia references deployments with retailers and travel brands, and third‑party profiles note clients including Qantas, Holland & Barrett and Starbucks along with a reported platform candidate‑experience score above 9/10 and multi‑million interview totals. (Sapia.ai; G2). U.S. case law and settlements have already tested AI hiring tools: a federal court preliminarily certified a collective action in the Mobley v. Workday matter in 2025, and CVS settled an AI‑interview related suit in July 2024, signaling legal exposure for automated screening outcomes. (Davis Wright Tremaine; HR Dive). Boards and retained search firms are adopting AI matching and screening tools: Boardsi published an overview of “AI‑powered board matching” in 2025, and major search firms such as Korn Ferry have described growing AI‑enabled talent platforms alongside traditional retained search services. (Boardsi; Korn Ferry). Sapia’s own fairness research and whitepapers describe an “auditable” approach and recommend checks on model outputs and human review in hiring flows — a governance step that aligns with regulatory transparency pushes such as requirements discussed in analyses of the EU AI Act and recent U.S. litigation. (Sapia.ai whitepaper; NatLawReview).
