Using AI in Recruitment

This week, the UK's Information Commissioner's Office (ICO) released its "Recruitment Rewired" report (see link in comments), and it’s a massive wake-up call for our industry here in the UK. If your Talent Acquisition team uses AI to sift, score, or rank candidates, the regulatory playbook has officially changed. For years, the industry has leaned on the "Human-in-the-Loop" defence to avoid the strict rules of Automated Decision-Making (ADM). The ICO has just closed that loophole. Here is the TL;DR on what HR and TA leaders need to know right now: ➡ 𝗧𝗵𝗲 "𝗥𝘂𝗯𝗯𝗲𝗿-𝗦��𝗮𝗺𝗽" 𝗜𝗹𝗹𝘂𝘀𝗶𝗼𝗻 𝗶𝘀 𝗗𝗲𝗮𝗱 • If an AI gives a candidate a "Red" fit score and a hiring manager simply clicks 'reject' without a meaningful review of that specific application, the ICO classifies this as a solely automated decision. Token human involvement no longer protects you from ADM regulations. ➡ 𝗧𝗵𝗲 𝗣𝗶𝘃𝗼𝘁 𝘁𝗼 𝗟𝗲𝗴𝗶𝘁𝗶𝗺𝗮𝘁𝗲 𝗜𝗻𝘁𝗲𝗿𝗲𝘀𝘁𝘀: • Thanks to the Data (Use and Access) Act 2025, there is a clearer path forward. The ICO explicitly advises moving away from "Consent" (which is rarely 'freely given' by desperate job seekers) and instead relying on "Legitimate Interests" to process high-volume AI hiring. ➡ 𝗧𝗵𝗲 𝗔𝗿𝘁𝗶𝗰𝗹𝗲 𝟮𝟮𝗖 "𝗖𝗮𝘁𝗰𝗵" • You are allowed to use fully automated sifting, but you must implement strict safeguards. If an AI rejects a candidate, you must explain the logic, give them the right to contest the decision, and offer the right to request a manual human review. ➡ 𝗩𝗲𝗻𝗱𝗼𝗿 𝗕𝗶𝗮𝘀 𝗶𝘀 𝗡𝗼𝘄 𝗬𝗼𝘂𝗿 𝗟𝗶𝗮𝗯𝗶𝗹𝗶𝘁𝘆 • You can no longer hide behind your tech provider if an algorithm discriminates. The ICO expects employers to demand bias testing results during procurement, run their own independent fairness trials, and actively monitor recruitment outcomes for ongoing bias. Ignorance of the algorithm is not a legal defence. ➡ 𝗗𝗣𝗜𝗔𝘀 𝗮𝗿𝗲 𝗡𝗼𝗻-𝗡𝗲𝗴𝗼𝘁𝗶𝗮𝗯𝗹𝗲 • Pointing to your tech vendor’s privacy policy is no longer sufficient. Employers are accountable, and if your Data Protection Impact Assessment doesn't explicitly map out these new AI safeguards and bias mitigations, you are exposed. 𝗛𝗲𝗿𝗲'𝘀 𝗺𝘆 𝘁𝗮𝗸𝗲: The ICO is not trying to kill volume hiring or innovation. They are giving the industry permission to automate, but demanding absolute transparency and fairness in return. More on this soon...

SmartRecruiters : the features that matter (and why teams adopt it fast) If you’re evaluating or implementing SmartRecruiters, here’s the feature map I use to align recruiters + hiring managers: ✅ ATS (end-to-end hiring workflow) Track applicants, standardize stages, keep everyone in one place (no “where is the latest resume?”). ✅ CRM (talent pools + nurture) Build pipelines for future roles and run outreach campaigns with measurable engagement. ✅ AI support (matching + assistant capabilities) Matching helps prioritize screening; AI-assistant concepts focus on removing repetitive work (screening/scheduling/engagement). ✅ Job Distribution Publish jobs broadly from one place (key for volume hiring and consistency). ✅ Dynamic Scheduling Reduce the back-and-forth for interviews (often one of the biggest time sinks). ✅ Messaging (SMS) Faster response loops with candidates—especially for high-volume roles. ✅ Integrations & Marketplace ecosystem Pre-built ecosystem to connect tools across sourcing, assessments, HRIS, etc. ✅ LinkedIn integrations (RSC + reporting) Sync candidate/app data with LinkedIn Recruiter and get ATS-enabled reporting (depending on your setup/contracts). My practical takeaway: SmartRecruiters is strongest when you design it as a hiring operating system (process + adoption), not just “an ATS implementation.” #SmartRecruiters #SAPSuccessFactors #ATS #Recruitment #TalentAcquisition #HRTech

Amazon’s hiring AI once rejected qualified women and preferred men. Here’s why: Paola Cecchi-Dimeglio, a Harvard lawyer and Fortune 500 advisor, has a warning for HR: If you ignore AI bias, you scale discrimination because it learns our prejudice and amplifies it in hiring and performance decisions. Remember Amazon's hiring algorithm? It systematically favored male candidates because it learned from historical hiring data that was already biased. The tool was discontinued, but the lesson remains relevant for every organization using AI today. Dimeglio identifies three critical sources of bias: 1. Training data bias: When AI learns from unrepresentative data, it produces skewed outcomes. For example, generative AI models underrepresent women in high-performing roles and overrepresent darker-skinned individuals in low-wage positions. 2. Algorithmic bias: Flawed data leads to biased algorithms. Recruitment tools may favor keywords more common on male resumes, perpetuating gender disparities in hiring. 3. Cognitive bias: Developers' unconscious biases influence how data is selected and weighted, embedding prejudice into the system itself. Paola's solution framework for HR leaders: ✅ Ensure diverse training data – Invest in representative datasets and synthetic data techniques  ✅ Demand transparency – Require clear documentation and regular audits of AI systems  ✅ Implement governance – Establish policies for responsible AI development  ✅ Maintain human oversight – Integrate human review in AI decision-making  ✅ Prioritize fairness – Use methods like counterfactual fairness to ensure equitable outcomes  ✅ Stay compliant – Follow regulations like the EU's AI Act and NIST guidelines As Paola emphasizes: "HR leaders, as the gatekeepers of talent and culture, must take the lead on avoiding and mitigating AI biases at work." This isn't just about fairness, it's about achieving better outcomes, building trust, and protecting your organization from legal and reputational risks. The question isn't whether AI has bias. It's whether you're doing something about it. How is your organization addressing AI bias in HR processes? Let's discuss.

The old way: Manual screening of thousands of CVs. The new way: #Agentforce. Capita's contact centre job listings attract tens of thousands of applications. Customers need those centres staffed up fast. But manual workflows have slowed the process, impacting candidates and customers. That’s why Capita's recruitment-as-a-service will use Salesforce Agentforce #AI agents to automate candidate matching and engagement. So they can help their customers fill business-critical roles – fast. Agentforce will help Capita quickly transform the recruitment process by autonomously taking action on early-stage tasks, such as enabling candidates to find jobs that fit their needs, assessing thousands of CVs in seconds, and narrowing the candidate pool for a potential match. For example, a recent graduate might come to Capita’s website looking for a position. Agentforce will ask what they’re looking for, prompt them to upload their CV, instantly analyse it, and suggest relevant roles. Once they apply, Agentforce can then suggest next steps for the human recruiter, helping them move qualified candidates through the hiring process faster — a significant advantage for businesses that need to keep thousands of roles filled or staff up quickly for holiday seasons and peak campaigns. Read their story: https://lnkd.in/eZpjbfS9

Efficiency has never been a moral defense for poor hiring, yet we are seeing a massive ripple effect across the industry following the class action lawsuit against Eightfold AI. The core of the issue is not just about whether AI works. It is about the enriched talent profiles created behind the scenes through data and inferences that candidates did not provide and cannot even see. In our desperate rush to save time and filter through the noise of high applicant volumes, we have moved from assistive tools to decisive systems that even the recruiters using them do not fully understand. We have seen this movie before with informal social media background checks, but today’s tech does not just replicate those instincts; it industrializes them. The hard truth is that if you cannot explain how a hiring decision was reached, you should not be making it. This lawsuit serves as a loud reminder that hiring is not ad targeting and candidates are not just datasets to be manipulated by an opaque algorithm. We need to stop hiding behind the "the system flagged them" excuse and start demanding a higher bar for transparency and accountability. Innovation is vital, but it must support human judgment rather than bypass it. We do not just choose our HR tech; we choose our standards. It is time we chose clarity over shortcuts. Read the full breakdown in my latest LinkedIn newsletter.

300 resumes for one role and your best candidate just ghosted you after waiting three weeks for feedback. This scenario plays out daily in recruiting teams everywhere. AI Recruiting Agents offer a different path forward. Think beyond the hype for a moment. These agents handle the repetitive tasks that drain your team's energy: resume screening, candidate ranking, interview scheduling, skill test deployment. All automated. What fascinates me is how they learn. Every hiring decision becomes training data. They recognize patterns, spot which traits predict success in your organization, and identify potential beyond the resume. The integration piece matters too. They plug into tools you already use while your recruiters focus on what humans do best: building relationships, reading between the lines, and making nuanced judgment calls. The data tells the story: 35% faster time-to-hire and 20% higher candidate satisfaction for companies using AI in 2024. That's competitive advantage. Of course, bias remains a real concern. Unchecked AI can perpetuate hiring mistakes from the past. Building in transparency and audit trails isn't negotiable. How are you balancing speed with quality in your hiring process right now? My thoughts on this are below in the comments. #recruitment #recruiting #hiring #HR #HumanCapitalist

I’m observing a deeply concerning trend in the current hiring environment. Some companies are inviting candidates to video interviews with their camera and microphone on, claiming to assess communication, reasoning, and problem-solving. Candidates are asked to participate in full interview rounds, complete realistic tasks and case studies, demonstrate thought processes & behavioural responses and even sign NDAs beforehand. However, the purpose is not to hire. The interviews are recorded and used to train multiple AI models, from conversational agents to behaviour and speech analysis systems. No real role. No real consideration. Just the extraction of human intelligence, emotional labour, and lived experience repackaged as “data”. This is exploitation disguised as innovation. Responsible AI must involve transparency of intent, explicit consent, not buried in legal language, fair compensation for the human effort and expertise provided, respect for candidates as people, not training material. If the industry normalizes this behaviour, we risk undermining trust and compromising the dignity of work itself. We must set higher ethical standards and hold each other accountable.

AI recruiting used to be a complete black box. Models were trained on mountains of data, then spat out answers with zero explanation. No visibility into why. No control over the output. LLMs have changed the game entirely. Now with Gem‎, when our AI ranks candidates, it doesn't just give you a match score – it tells you exactly WHY that candidate earned that score: - What specific aspects of their background led to the rating? - What criteria were met? When something's off, recruiters can adjust the criteria and get better matches next time. This explainability helps reduce bias, too. When AI is a black box, you have no idea if underlying biases are influencing results. With transparent reasoning, you can identify and eliminate those issues. Steve DeCorpo, Director of Global Talent Acquisition (Celestica), calls Gem's ability to narrow down and rank large numbers of applications with a click "a game changer" for identifying perfect candidates. Katie Durvin, Senior Recruitment Manager (Fingerprint), found that inputting job requirements resulted in applicants being scored perfectly, showing how well our AI aligns with recruiter expertise. That's why we're not trying to replace recruiters with AI. We're putting recruiters firmly in the driver's seat, creating an iterative loop where human expertise and AI capabilities enhance each other. The recruiter defines criteria, the AI explains its reasoning, the recruiter refines the approach, and the process improves with each cycle. Control. Visibility. Collaboration. That's the evolution of AI in recruiting.

Data privacy and ethics must be a part of data strategies to set up for AI. Alignment and transparency are the most effective solutions. Both must be part of product design from day 1. Myths: Customers won’t share data if we’re transparent about how we gather it, and aligning with customer intent means less revenue. Instacart customers search for milk and see an ad for milk. Ads are more effective when they are closer to a customer’s intent to buy. Instacart charges more, so the app isn’t flooded with ads. SAP added a data gathering opt-in clause to its contracts. Over 25,000 customers opted in. The anonymized data trained models that improved the platform’s features. Customers benefit, and SAP attracts new customers with AI-supported features. I’ve seen the benefits first-hand working on data and AI products. I use a recruiting app project as an example in my courses. We gathered data about the resumes recruiters selected for phone interviews and those they rejected. Rerunning the matching after 5 select/reject examples made immediate improvements to the candidate ranking results. They asked for more transparency into the terms used for matching, and we showed them everything. We introduced the ability to reject terms or add their own. The 2nd pass matches improved dramatically. We got training data to make the models better out of the box, and they were able to find high-quality candidates faster. Alignment and transparency are core tenets of data strategy and are the foundations of an ethical AI strategy. #DataStrategy #AIStrategy #DataScience #Ethics #DataEngineering

AI can cut hiring time by 80% (McKinsey & Company), but at what cost? Automation is faster, smarter, more efficient, but if we’re not careful, it’s also more biased, less human, and dangerously flawed. As a result, HR leaders now hold a double-edged sword. + Use AI wisely, and it transforms recruitment.  + Use it blindly, and it reinforces the very problems we’re trying to solve. According to McKinsey, AI-driven tools have increased recruiting efficiency by 80%, yet 76% of job seekers say the hiring experience impacts whether they accept an offer. Speed matters.  But so does fairness.  So does trust. Because efficiency means nothing if candidates feel reduced to a data point. AI is only as fair as the data it learns from. And if that data carries bias? AI will replicate it, at scale. I still remember an instance from two years back: a candidate with an unconventional career path, a late-degree switch, a few gaps, non-traditional experience was filtered out by an AI-automated software. On paper, they weren’t a fit. In reality, they were exactly what the company needed. But imagine how many great hires are being lost because no one is watching? AI can analyse resumes, predict job fit, and streamline hiring like never before. But it cannot replace the human judgment, emotional intelligence, and ethical responsibility that recruiters bring to the table. So, how do we use AI without losing the human element? ✅ Train AI to spot bias, not amplify it: AI learns from past data. If that data carries bias, AI will replicate it. Audit algorithms. Diversify data sets. Ensure AI isn’t just fast, but fair. ✅ Use AI to enhance decision-making, not replace it: Predictive analytics can tell you who to interview. But only humans can assess cultural fit, build trust, and make final hiring decisions. ✅ Create transparency in hiring: Candidates should know when AI is evaluating them. If an algorithm rejects someone, recruiters should intervene, not blindly trust the machine. ✅ Prioritise candidate experience: Chatbots and automation can provide instant updates, but real conversations build relationships. The best hires don’t just want a job, they want to feel valued. AI isn’t the future of recruitment. Humans + AI is. The goal isn’t to replace recruiters, it’s to empower them to be better, faster, and fairer. Because at the end of the day, great hiring isn’t just about efficiency. It’s about people. #aiinhr #ethicalhiring #hrleadership Puneet Chandok, Navnit Singh, Rishi Khandelwal, Shailja Dutt