AI Can Fake Identity. It Can't Fake Humanity

AI Governance is a relationship, not a document

Summary On The Human Element, Barb Bidan sits down with Joshua Horenstein, who runs both the HR and legal functions at Innophos. Joshua argues that AI governance fails on the org chart, not in the policy document. He walks through his manufacturing-pilot mindset for rolling out AI, names the two reasonable worst cases most HR leaders aren't watching for, and explains why the friction between the CHRO and the GC is the engine of good policy, not a problem to fix. For HR leaders, CHROs, and the chief legal officers who work alongside them. Chapters 00:00 Introduction 00:45 Joshua's dual hat: CHRO and chief legal officer at Innophos 03:15 The first governance question, what type of AI are you deploying 04:45 The pilot-to-policy mindset from manufacturing 05:45 AI governance as the rules of the road 07:30 Two reasonable worst cases: productivity backfire and dormant data leak 10:45 Sandboxing AI in an industrial company 13:45 The CHRO and CLO tension as the policy engine 16:45 Curiosity beats the playbook in HR leadership 23:15 AI as your shadow partner, the two-year forecast 26:15 The biggest flaw in current AI governance models Takeaways * AI governance is the rules of the road, not a 50-subpart document. Flexible top-level rules outlast rigid policies that age before they're signed. * The internal AI data leak (a dormant folder permission surfaced by an AI tool) is almost as dangerous as the external one, and almost nobody is auditing for it. * The friction between CHRO and CLO is the engine of good AI policy. Siloed functions produce one-sided answers. * Sandbox the AI like a chemical pilot. Cordon off the data, open the spigots gradually, watch for leaks. * Never let AI make a termination, promotion, or compensation decision. Humans own the outcome. Connect with the Guest LinkedIn: https://www.linkedin.com/in/joshhorenstein/ [https://www.linkedin.com/in/joshhorenstein/] Website: https://innophos.com [https://innophos.com] Sponsor Wisq is the AI platform for HR. We built Harper, the world's first AI HR teammate, designed to handle the judgment-heavy work that has historically consumed HR teams: job changes, performance concerns, leaves of absence, onboarding, and employee relations issues. Where HCM bolt-ons and chatbots deflect the easiest questions, Harper resolves full cases, end-to-end. The result is an HR function with the capacity and strategic bandwidth to focus on the work that moves the business. Companies get Harper live in weeks, not quarters, with implementation support built on deep HR domain expertise. Wisq serves HR leaders at leading companies across industries, helping them raise the bar on employee experience and expand what their teams are capable of. For more information, visit https://www.wisq.com [https://www.wisq.com] * (00:00) - Introduction * (00:45) - Joshua's dual hat: CHRO and chief legal officer at Innophos * (03:15) - The first governance question, what type of AI are you deploying * (04:45) - The pilot-to-policy mindset from manufacturing * (05:45) - AI governance as the rules of the road * (07:30) - Two reasonable worst cases: productivity backfire and dormant data leak * (10:45) - Sandboxing AI in an industrial company * (13:45) - The CHRO and CLO tension as the policy engine * (16:45) - Curiosity beats the playbook in HR leadership * (23:15) - AI as your shadow partner, the two-year forecast * (26:15) - The biggest flaw in current AI governance models

AI Can Fake Identity. It Can't Fake Humanity

Description In December 2024, Christine Aldrich's recruiting team caught their first deepfake candidate. Twelve months later, nearly 1 in 6 applicants at Pindrop shows signs of data manipulation — and every HR leader Christine knows has a story. In this episode of The Human Element, host Barb Bidan sits down with Christine Aldrich, Chief People Officer at Pindrop (the company building identity verification and deepfake detection technology), to talk about what hiring really looks like when you can no longer assume the person on the other end of the camera is who they say they are. They cover the velocity of the deepfake hiring problem, why the two reflex responses (ban AI / go back in-person) are both wrong, the lightweight tactic any recruiting team can deploy on Monday morning, and why the future of recruiting belongs to the people who let technology verify identity so humans can focus on what only humans can do. Takeaways * Why December 2024 was the inflection point for deepfake candidates in hiring * The Pindrop data behind the "1 in 6" stat — and what it really means for your funnel * Why banning AI in your application process filters out your best candidates while letting fraudsters through * How recruiter bias training collides with deepfake detection — and how to resolve the conflict * The three-minute consistency check any recruiter can start using on Monday * How to run identity verification in the background so candidate experience doesn't suffer * Why "AI fluency" is now a behavioral attribute at Pindrop * The most overrated tool in fighting candidate fraud (hint: you've been on it today) * What recruiting looks like when AI takes the bottom of the skill stack Christine's closing reframe for every HR leader on the drive home Chapters (00:00) Cold open and intro (01:15) Christine's path: Michigan, Ford, Dunkin' Brands, and four startups (04:00) The first fraudulent candidate (December 2024) and the velocity of the problem (05:45) The 1-in-6 stat and what data manipulation actually looks like (06:30) What a deepfake candidate looks like in a real interview (09:00) AI usage vs. AI fabrication — where Pindrop draws the line (11:00) Why this isn't just a tech industry problem (12:30) The collision between recruiter bias training and fraud detection (16:00) The lightweight tactics any team can start using Monday morning (18:00) Verification in the background — preserving candidate experience (21:30) AI as both threat and defense — and why treating it as only one is the mistake (23:30) How Pindrop's own team uses AI in the recruiting process (25:30) The candidate side of the verification conversation (28:00) Lightning round: first red flag, the recruiter tactic, the most overrated tool (29:30) "AI can fake identity. It can't fake humanity." (31:00) Barb's takeaways and Christine's final word Connect with the Guest * LinkedIn: https://www.linkedin.com/in/chriskaszubski/ [https://www.linkedin.com/in/chriskaszubski/] * Pindrop: https://www.pindrop.com [https://www.pindrop.com] Sponsor Wisq is the AI platform for HR. We built Harper, the world's first AI HR teammate — designed to handle the judgment-heavy work that has historically consumed HR teams: job changes, performance concerns, leaves of absence, onboarding, and employee relations issues. Where HCM bolt-ons and chatbots deflect the easiest questions, Harper resolves full cases, end-to-end. The result is an HR function with the capacity and strategic bandwidth to focus on the work that moves the business. Companies get Harper live in weeks, not quarters, with implementation support built on deep HR domain expertise. Wisq serves HR leaders at leading companies across industries, helping them raise the bar on employee experience and expand what their teams are capable of. For more information, visit  https://www.wisq.com [https://www.wisq.com/]

How to Onboard Your AI the Right Way

Summary On The Human Element, Barb Bidan talks with Q Hamirani, Chief People Officer at HighLevel, about treating AI as a teammate you onboard rather than a tool you install. Q draws on building people functions at Airbnb, Paper, and now a 2,000-person fully remote AI company to argue that the real constraint on AI adoption is human cognitive capacity, not technology. He shares how he trained an AI compensation model to stop and validate at three checkpoints so a human always owns the outcome. The throughline: the people leader's job is not to implement AI, it is to decide what stays human. Built for CHROs, VPs of People, and anyone leading AI adoption in HR. Chapters 00:00 Meet Q Hamirani, CPO at HighLevel 02:20 Backing into HR from an engineering mindset 06:10 AI as a teammate, not a tool 11:00 Defining what a digital teammate owns 13:05 The real bottleneck is human capacity 16:40 Multi-modality: surveys, songs, and websites 21:00 Three stops that keep AI out of the black box 24:40 Accountability as culture, not compliance 27:30 Bolting AI onto broken work 30:40 Lightning round and the last word Takeaways - Onboard AI like a teammate with a loose job description and a clear owner, not a tool you install. - The binding constraint on AI adoption is human cognitive capacity, not budget or technology. - Break AI processes into blocks with human validation stops so nothing runs as an end-to-end black box. - Accountability does not change in an AI world; a human still owns the outcome. - The people leader's job is to decide what stays human. Connect with the Guest LinkedIn: https://www.linkedin.com/in/hamirani/ [https://www.linkedin.com/in/hamirani/] Website: https://www.gohighlevel.com/ [https://www.gohighlevel.com/] Sponsor Wisq is the AI platform for HR. We built Harper, the world's first AI HR teammate — designed to handle the judgment-heavy work that has historically consumed HR teams: job changes, performance concerns, leaves of absence, onboarding, and employee relations issues. Where HCM bolt-ons and chatbots deflect the easiest questions, Harper resolves full cases, end-to-end. The result is an HR function with the capacity and strategic bandwidth to focus on the work that moves the business. Companies get Harper live in weeks, not quarters, with implementation support built on deep HR domain expertise. Wisq serves HR leaders at leading companies across industries, helping them raise the bar on employee experience and expand what their teams are capable of. For more information, visit wisq.com [http://wisq.com/].

When Doing Great HR Work Isn't Enough

Summary In this episode of The Human Element, Chief People Officer John Foster (Tala) makes the case that the biggest risk for HR leaders isn't doing bad work — it's doing great work that the organization isn't ready to receive. John shares the framework he calls scratch OD, his argument for why AI won't fix performance management but will replace it, and the career moment when a boss told him to take his foot off his throat. Essential listening for people leaders who want to use the AI era to redesign, not just optimize. Chapters 00:00 Introduction: John Foster, Tala, and the Gamut framework 02:30 AI as restructuring work, not optimizing it 05:30 Organization market fit: the Cambrian explosion analogy 09:30 Applying AI at Tala: rethinking software development and workflows 11:30 Performance management's 10-second reinvention 15:30 Using AI as a thinking partner for organizational rethinking 19:00 Scratch OD: the chef framework for HR design 23:00 The change agent's trap: empathy for overwhelmed leaders 26:30 When HR is correct but not effective 28:00 The winning organization in 2026: customer-centric and modular 30:00 Lightning round: Trader Joe's, Ford, and what HR most often skips Takeaways 1. Being entirely right about the work does not make you effective at changing the organization. 2. Scratch OD starts with the ingredients — what people actually need from work — not with which best practice to borrow. 3. AI makes individual performance clarity at scale possible for the first time, shifting the model from top-down judgment to individual empowerment. 4. The organizations that thrive in the AI era will be customer-centric and modular, not defined by their product or technology. 5. If you can imagine what you want to build and describe it in detail, you already have an entry point into using AI to construct it. Connect with the Guest LinkedIn: https://www.linkedin.com/in/johnfoster-gamut/ [https://www.linkedin.com/in/johnfoster-gamut/] Website: https://tala.co/careers/ [https://tala.co/careers/] Sponsor Wisq is the AI platform for HR. We built Harper, the world's first AI HR teammate — designed to handle the judgment-heavy work that has historically consumed HR teams: job changes, performance concerns, leaves of absence, onboarding, and employee relations issues. Where HCM bolt-ons and chatbots deflect the easiest questions, Harper resolves full cases, end-to-end. The result is an HR function with the capacity and strategic bandwidth to focus on the work that moves the business. Companies get Harper live in weeks, not quarters, with implementation support built on deep HR domain expertise. Wisq serves HR leaders at leading companies across industries, helping them raise the bar on employee experience and expand what their teams are capable of. For more information, visit wisq.com [http://wisq.com]. * (00:00) - Introduction: John Foster, Tala, and the Gamut framework * (02:30) - AI as restructuring work, not optimizing it * (05:30) - Organization market fit: the Cambrian explosion analogy * (09:30) - Applying AI at Tala: rethinking software development and workflows * (11:30) - Performance management's 10-second reinvention * (15:30) - Using AI as a thinking partner for organizational rethinking * (19:00) - Scratch OD: the chef framework for HR design * (23:00) - The change agent's trap: empathy for overwhelmed leaders * (26:30) - When HR is correct but not effective * (28:00) - The winning organization in 2026: customer-centric and modular * (30:00) - Lightning round: Trader Joe's, Ford, and what HR most often skips

The Moments That Make or Break Your HR Function

Summary In this episode of The Human Element, Barb talks with Heather Oxley, CHRO at Perficient, about what it looks like to run HR for a firm whose entire business is helping other companies transform. Heather walks through three areas where she's deploying AI across the talent life cycle: project staffing, learning and development, and recruiting. Her central argument is that the value of AI in HR isn't primarily about cost reduction — it's about bringing clarity to the decisions that define the function. She also names the one thing AI should never touch. Chapters 00:00 Host Intro: HR at a Firm Built on Transformation 01:30 Heather's Role at Perficient and the AI-First Mandate 03:30 AI in Project Staffing: Beyond Skills and Availability 09:00 The Leapfrog Moment for Smaller Teams 12:30 Learning as Career Navigation, Not Content Delivery 17:00 90% Voluntary Adoption in Two Weeks 19:30 AI in Recruiting: Where the Legal Stakes Are Highest 22:30 Closing the Candidate Black Hole and Debiasing Panels 24:30 The Fork in the Road: Cost Reduction vs. Decision Clarity 26:00 Lightning Round: The Line AI Should Never Cross Takeaways 1. AI in recruiting should augment human judgment, not replace it, especially as legislation in New York, California, and Colorado continues to evolve. 2. Using agents to close the loop with every candidate who doesn't advance is one of the highest-value, lowest-risk uses of AI in HR. 3. The fork facing HR leaders today is cost reduction vs. decision clarity, and most organizations are defaulting to cost. 4. Moments that matter (offers, terminations, promotions, hard conversations) should always involve a person, not an agent. 5. The HR leaders who matter in this era are the ones who prove the function can be as agentic as any other part of the business. Connect with the Guest LinkedIn: https://www.linkedin.com/in/heatheroxley/ [https://www.linkedin.com/in/heatheroxley/] Website: https://www.perficient.com/ [https://www.perficient.com/] Sponsor The Human Element is brought to you by Harper, Wisq's always-on AI HR Generalist transforming how work gets done. Powered by deep HR intelligence, Harper delivers instant, accurate, and empathetic support—from policy questions to performance coaching—so your people get answers fast and your teams stay focused. Learn more at https://www.wisq.com/ [https://www.wisq.com/] * (00:00) - Host Intro: HR at a Firm Built on Transformation * (01:30) - Heather's Role at Perficient and the AI-First Mandate * (03:30) - AI in Project Staffing: Beyond Skills and Availability * (09:00) - The Leapfrog Moment for Smaller Teams * (12:30) - Learning as Career Navigation, Not Content Delivery * (17:00) - 90% Voluntary Adoption in Two Weeks * (19:30) - AI in Recruiting: Where the Legal Stakes Are Highest * (22:30) - Closing the Candidate Black Hole and Debiasing Panels * (24:30) - The Fork in the Road: Cost Reduction vs. Decision Clarity * (26:00) - Lightning Round: The Line AI Should Never Cross