AI Hiring: Candidate Insights and Challenges

I recently interviewed with an AI agent. No human on the other side. Just me, a screen, and a voice asking questions and scoring my answers while I was talking. Honestly, it was less awkward than I expected. Fast, easy, strange maybe. Not my favourite type of interaction, but that is not really the point. It made me wonder: If this is how we screen or interview people now, how do we actually know it works? Then I read this: “Only 20% of organisations track quality of hire.”— SHRM, 2025 Benchmarking Survey. We are building all these fancy tools to automate the parts of recruitment that take a lot of time. Screening, scheduling, first interviews etc. And yes, that part is great. It saves hours. It can also affect the candidate experience, but that is another conversation. What we are automating is the process, but how about the outcome? Most companies still don’t properly measure whether the people they hire actually become good hires. 1. How do we automate the measuring part too, not just the screening? 2. How do we know if these AI agents improve quality of hire, or if they just speed up the process? 3. And will AI finally give us more consistent hiring data, or are we moving faster without really knowing where we are going? I genuinely don’t have the answer. I am asking because I am curious. If your company uses AI in hiring, do you track quality of hire? And how?

A lot of companies now tell candidates: “Do not use AI during the application or interview process.” At the same time, those same companies are using AI to scan CVs, rank applicants, analyse interview responses, score coding tests, monitor behaviour patterns, and sometimes even decide who gets filtered out before a human ever looks at them. That feels like a very strange imbalance. If AI is considered an unfair advantage for candidates, why is it acceptable for employers to use it across the entire hiring funnel? The reality is AI is already part of modern work. Most jobs today expect people to use copilots, automation, search, summarisation, or AI-assisted workflows once they are hired. So banning candidates from using the same tools during hiring feels disconnected from the actual workplace. Maybe the better question is not “Did you use AI?” Maybe it should be: Can you think critically? Can you validate outputs? Can you communicate clearly? Can you use AI effectively without losing human judgement? Because that is probably the real skill companies will need going forward. Curious where others stand on this. Is this fairness, necessary control, or just corporate hypocrisy dressed up as policy?

The "AI Narcissism" Trap in 2026 Hiring As we move into the era of agentic hiring, we need to be mindful of AI Self-Preference Bias. When you use an LLM to screen resumes, it doesn't just look for qualifications, it looks for its own reflection. The 2026 Benchmarks: 80%+ Bias: Major models (GPT, Claude, LLaMA) are significantly more likely to shortlist resumes they "wrote" themselves over equally qualified human versions. The Cross-Model Penalty: Using a different AI for writing vs. screening can result in a "style mismatch" that gets top-tier candidates rejected. Signal Loss: In a world of "perfect" AI resumes, the document has lost its ability to prove competence. The Strategic Reality: Recruiters are no longer just "sourcers"; they are Authenticators. In 2026, the recruiter’s edge is verifying Substance over Syntax. While AI is busy rewarding its own linguistic patterns, a human recruiter can verify if that "perfect" candidate actually fits your mission and culture and identify signals to best sell to. This is especially important when hiring for start-ups. Is your screening tool finding the best talent, or just the best-prompted resume? [sources in comments] P.S. you can use AI to mitigate bias in the process tools like Metaview allow you analyse interviewer performance, including that of recruiters.

How Different AI Tools Can Help Recruiters 𝗕𝗲𝗰𝗼𝗺𝗲 𝗯𝗲𝘁𝘁𝗲𝗿 𝗮𝘁 𝗔𝗜 𝗶𝗻 𝗷𝘂𝘀𝘁 𝟭 𝗺𝗶𝗻𝘂𝘁𝗲 𝗮 𝗱𝗮𝘆. 𝗚𝗲𝘁 𝘁𝗵𝗲 𝗔𝗜 𝗻𝗲𝘄𝘀𝗹𝗲𝘁𝘁𝗲𝗿 𝘀𝗺𝗮𝗿𝘁 𝗹𝗲𝗮𝗱𝗲𝗿𝘀 𝗿𝗲𝗮𝗱. 𝗦𝗶𝗴𝗻 𝘂𝗽 𝗳𝗿𝗲𝗲 𝗻𝗼𝘄 → aiforleaders.com Original post: __________ Save this post. Every recruiter using AI right now needs this breakdown. There are 4 AI tools that work best in recruiting but only when you use each one for what it's actually good at: 🟢ChatGPT — Writing & Structuring Use it to: → Write candidate outreach messages → Generate interview questions → Create job board versions → Improve JD structure → Separate must-have and nice-to-have skills → Rewrite vague responsibilities → Remove biased wording → Write clear, compelling job descriptions 🟠Claude — Review & Refinement Use it to: → Turn rough notes into polished drafts → Improve section structure and flow → Remove unrealistic requirements → Simplify complex language → Check candidate fit against the JD → Find missing role details → Improve tone and review JD clarity 🔵Gemini — Research & Market Context Use it to: → Research similar roles and common skills → Suggest relevant keywords → Compare similar JDs in the market → Track hiring trends → Create seniority-based versions → Benchmark responsibilities → Improve search visibility ⚫Catch.hr — AI Hiring Execution This is where everything connects: → Match roles with the right candidates → Source active and passive talent → Screen candidates early → Filter low-fit profiles automatically → Enrich candidate data → Automate outreach and follow-ups → Create cleaner shortlists → Move faster from JD to hire ChatGPT writes it. Claude refines it. Gemini researches it. Catch.hr executes it. That's how the fastest hiring teams in 2026 operate. 👉Which of these 4 tools are you already using in your recruiting workflow? Drop it in the comments. 👉Try Catch.hr: https://www.catch.hr/ Credit to Marco Verhoeven. Follow him for more.

One of the clearest places I am seeing weak communication and lazy judgement show up right now is in hiring. And what blows my mind is how often I hear people discredit or disregard an application because a candidate may have used AI in the process. When in fact, the business is using AI too! To write the ad, screen the CV, summarise the candidate, craft the interview questions. Recent hiring data says 65% of employers are already using AI in hiring, while only 37% of job seekers say they use AI to complete applications. Businesses are building AI into the process, then clutching their pearls because a candidate may have used AI too. Talk about hypocrisy! And yes, there is a difference between using AI to support a process and using it to fake substance. And that is exactly my point. The real question is not, did they use AI? It is, can the human underneath it still think, communicate and show judgement when it counts? Because the real signal is not in the scripted question. It is what you do with the answer. It is hearing something and knowing where to dig... "earlier you mentioned this and I want to go back to that." It is pulling on the thread and asking the follow-up that tells you how someone actually thinks, reflects, communicates and responds when things move off script. That is where you find out if someone has depth or just a polished answer. And right now, too much of that signal is weak. Candidates are showing up ready for predictable questions with surface-level answers. Interviewers are not much better. Basic questions in. Basic answers out. Very little depth. Very little substance. And the minute you ask someone to think harder, go deeper, give you a real example, explain the outcome, or unpack what actually happened, the wheels start wobbling. That is the part that concerns me. It is starting to feel like people are losing the ability to think, reflect and communicate with any real depth once they are pushed beyond the script. That matters because employers keep saying they want communication, problem-solving, and judgement, while also saying one of the hardest parts of hiring is actually working out whether candidates truly have the right soft skills. So if the process is getting flatter, more scripted and more hollow, that is not a small issue. That is the issue! Not just in hiring. In business. Because when people start outsourcing the thinking, communication gets flatter, judgement gets weaker, and decision-making gets worse. AI has a place. But if people are using it to bypass the hard human work instead of strengthening it, do not act surprised when the output feels hollow. We do not just need better tools. We need humans showing up better in the room. If this is showing up in your business, reach out. This is exactly the kind of problem Glass Partners helps solve. #Leadership #Communication #Hiring #AI #PeoplePerformance #BusinessPerformance #FutureOfWork #GlassPartners

💡 𝐅𝐮𝐭𝐮𝐫𝐞 𝐨𝐟 𝐇𝐢𝐫𝐢𝐧𝐠? A few years ago, consulting interviews were built around one core assumption: Put a candidate in front of a difficult business or technical problem, remove all external help, and observe how they think. And honestly, this made sense. The strongest candidates were usually the ones who could structure ambiguity, break complexity into smaller pieces, communicate clearly, and move logically under pressure. But I increasingly wonder whether this model is still testing the right thing for the future we are entering. Because consulting is changing faster than recruiting processes. Today, consultants, engineers, strategists, and analysts operate alongside AI systems constantly. AI copilots, coding agents, reasoning models, automated research assistants, simulation tools, and dynamic execution environments are quickly becoming part of daily work. So the question becomes: Why are we still interviewing people as if none of these tools exist? Recently, I saw discussions online about candidates secretly using AI during interviews. The immediate reaction was predictable: “This is cheating.” But maybe that reaction misses the bigger signal entirely. Maybe this is not exposing a candidate problem. Maybe it is exposing a recruiting problem. Because future high performers will not be the people who refuse to use AI. They will be the people who know how to orchestrate intelligence around them. The future consultant may look less like someone solving everything alone on a whiteboard, and more like a conductor coordinating humans, models, agents, tools, simulations, and live information streams in real time. So what if interviews were redesigned accordingly? Imagine candidates intentionally given access to multiple AI tools. Allowed to use coding agents. Allowed to build live prototypes. Allowed to automate workflows. Allowed to dynamically test ideas. Instead of testing who can solve everything alone under pressure, what if we started evaluating something entirely different: How candidates think out loud while collaborating with AI. How they prompt out loud. How they delegate tasks across AI systems. How they orchestrate models, tools, agents, and live information streams together. How they validate outputs instead of blindly trusting them. How they challenge weak reasoning. How they recover when AI produces flawed assumptions. How they turn raw ideas into working prototypes and executable solutions in real time. Because maybe the future of hiring is no longer about testing isolated intelligence. Maybe it is about testing amplified intelligence. And I genuinely wonder: Are Tier 1, Tier 2, Tier 3, and Tier 4 consulting firms already thinking this way? Because the firms that figure this out first may completely redefine what top talent looks like in the AI era.

Let’s talk about AI and recruiting for a hot minute…… The irony is that companies are criticizing candidates for using AI… while many companies are using AI to screen resumes before a human ever sees them. At this point, candidates almost have to use AI tools to optimize keywords and tailor resumes just to make it through ATS filters and reach a real person. That’s not “cheating” — that’s adapting to the hiring process companies created. There’s a huge difference between using AI to fabricate experience versus using it to improve formatting, clarity, keyword alignment, and communication. Resume writers and career coaches have been doing the same thing for years. The real issue isn’t whether AI assisted with the resume. It’s whether the candidate can actually back up their experience, skills, and results in an interview and on the job.

When I interviewed at Boston Consulting Group (BCG), there was no AI in the room. Just a case, an interviewer, and your ability to think under pressure. Today that same interview uses AI. In many countries, BCG uses Casey its AI chatbot that conducts your first round entirely. McKinsey & Company is going further. They're piloting Lilli (AI) inside final rounds where candidates solve real business problems with the AI live, while interviewers watch how you engage with it, challenge it, and own the final answer. The format has evolved and is evolving. But here's what hasn't: Firms aren't testing whether you can use AI. They're testing whether you can think when AI is in the room. While candidates are being blacklisted for outsourcing their reasoning to ChatGPT during virtual interviews. That's exactly what firms are screening against. AI fluency matters. But logic cannot be outsourced. While the process may have changed. The standard hasn't. #Consulting #MBB #Interview #AI #Career

People who are arguing about AI in hiring are arguing about the wrong thing. I'm actively working around this now, so if you're a candidate, let me hold your hand and break this down. The AI in hiring debate is now framed as: yes or no.  I really believe the framing is already obsolete. The moment a single job started pulling 100s/1000s of applications, the questions for businesses stopped being whether to use AI and became 𝘄𝗵𝗶𝗰𝗵 𝘀𝘁𝗮𝗴𝗲, 𝘄𝗵𝗶𝗰𝗵 𝗱𝗲𝗰𝗶𝘀𝗶𝗼𝗻, 𝗮𝗻𝗱 𝘄𝗶𝘁𝗵 𝗵𝗼𝘄 𝗺𝘂𝗰𝗵 𝗵𝘂𝗺𝗮𝗻 𝘄𝗲𝗶𝗴𝗵𝘁 𝗼𝗻 𝘁𝗼𝗽 𝗼𝗳 𝗶𝘁. The companies using these tools are not really seduced by them. They are using them because the alternative is a recruiter trying to screen enough people in an aggressively short timeline, and still missing people. Breaking news, they are missing on so many people out there, every single time. That doesn't make every implementation good. A lot of them are bad: - Some are measurably biased.  - Some are dressed-up screeners pretending to be interviews.  - Some are deployed by people who can't tell you what the AI is scoring. 𝗧𝗵𝗼𝘀𝗲 𝗮𝗿𝗲 𝗿𝗲𝗮𝗹 𝗽𝗿𝗼𝗯𝗹𝗲𝗺𝘀 𝗮𝗻𝗱 𝘁𝗵𝗲𝘆 𝗱𝗲𝘀𝗲𝗿𝘃𝗲 𝗿𝗲𝗮𝗹 𝗽𝗿𝗲𝘀𝘀𝘂𝗿𝗲. But I really think there is something candidates don't want to hear: 𝗿𝗲𝗳𝘂𝘀𝗶𝗻𝗴 𝗼𝗻 𝗽𝗿𝗶𝗻𝗰𝗶𝗽𝗹𝗲 𝗶𝘀 𝗮 𝗽𝗲𝗿𝘀𝗼𝗻𝗮𝗹 𝗲𝘅𝗶𝘁, 𝗮𝗻𝗱 𝘂𝗻𝗳𝗼𝗿𝘁𝘂𝗻𝗮𝘁𝗲𝗹𝘆, 𝘁𝗵𝗲 𝗰𝗼𝗺𝗽𝗮𝗻𝘆 𝗱𝗼𝗲𝘀𝗻'𝘁 𝗳𝗲𝗲𝗹 𝗶𝘁. The metrics a business will watch are things like: time-to-hire, cost-per-hire, funnel volume, etc. This doesn't change when you opt out. What moves these systems is internal: governance reviews, legal exposure, adverse impact audits, and candidates who engage and then give specific, structural feedback the company can't dismiss as bitterness. In the current cycle, the candidate with actual leverage right now: - The one who can tell the difference between an AI screener that's badly designed and one that's doing a defensible job. - It's the candidate who is asking what's being measured. - Its's the candidate who noticing when "AI interview" is really a one-way video with no human downstream, and call that out, by name, in writing. - It's the candidate who understand that a recruiter reviewing AI output is a very different thing from an AI making the call. That's the literacy that's going to matter for this 1-2 years period. Basically, the ability to read the system, locate the actual failure points, and why not push on those. You really have far more power in this than the loudest takes out there on LinkedIn. 𝗦𝗽𝗲𝗻𝗱 𝗶𝘁 𝘄𝗵𝗲𝗿𝗲 𝗶𝘁 𝗰𝗵𝗮𝗻𝗴𝗲𝘀 𝘀𝗼𝗺𝗲𝘁𝗵𝗶𝗻𝗴.

Is AI replacing human hiring judgment—or finally making hiring more objective? The conversation around AI in hiring often swings between hype and fear. But perhaps we’re asking the wrong question. The real challenge in software hiring today is not whether AI should participate—it’s whether humans alone can realistically evaluate talent at scale with consistency. A few realities in today’s hiring landscape: 🔹 The rise of fabricated CVs is real Keyword-optimized, AI-assisted, and sometimes outright misleading profiles are becoming increasingly common. Resume-based shortlisting alone is no longer enough. 🔹 Bias and oversight are unavoidable in manual screening Even experienced recruiters and hiring managers face fatigue, inconsistency, unconscious bias, and time constraints—especially when reviewing hundreds or thousands of profiles. 🔹 Talking to every candidate is ideal—but operationally impossible at scale The fairest hiring process would actually involve giving every potentially relevant candidate a chance to be evaluated beyond just what’s written on paper. But with traditional processes, that simply doesn’t scale. This is where reliable Agentic AI can fundamentally change hiring. Not generic automation. Not simplistic keyword filtering. But AI systems with checks, balances, contextual reasoning, and precision evaluation frameworks. The real value of AI in hiring is: - Consistent CV evaluation against role-specific criteria -Reduced human bias in early filtering -Ability to engage and assess far more candidates than human teams can -Structured technical evaluation beyond resume claims -Better signal-to-noise ratio for final human decisions At Caze Labs, this is exactly the problem we are solving with Caze HireSense. Our approach is intentionally two-stage: Stage 1: AI-driven CV evaluation using defined qualification thresholds and contextual scoring Stage 2: AI-led interview assessment for candidates who clear the initial benchmark This makes it possible to move beyond resume assumptions and actually evaluate candidates through interaction—at a scale impossible with purely manual hiring. The goal is not to remove humans from hiring. The goal is to ensure humans spend their time making decisions based on stronger, validated signals—not incomplete or fabricated inputs. AI should not replace judgment. AI should improve the quality of judgment. What’s your take—will trustworthy Agentic AI become a standard layer in software hiring? #AI #Hiring #Recruitment #TalentAcquisition #SoftwareHiring #AgenticAI #ArtificialIntelligence #HRTech #FutureOfWork #CazeHireSense #RecruitmentTechnology #EngineeringHiring