Who Do You Trust?

With public job data under fire and AI models failing basic competency tests, trust is more important than ever in talent acquisition.

This edition of the ERE Weekly is coming to you from the hills of Granada, Spain, where I am vacationing with my family. It’s absolutely beautiful here. And hot. Really, really hot. But hey, it’s a dry heat, right?

Don’t worry though—I’ve still got your weekly fix of what’s shaping the world of recruiting.

This week, I’ve been thinking a lot about trust. It’s critical to our relationships with leadership, hiring managers, and candidates. If any of these stakeholders lose faith in talent acquisition, it’s game over. We simply can’t do our jobs. In a sense, when tech bros tell us that AI is going to replace recruiting, what they are really saying is that they don’t trust TA professionals to do a better job than the AI models.  

Shooting the Messenger at the BLS

What got me thinking about trust was the Bureau of Labor Statistics debacle last week. The BLS released a disappointing July report (73,000 new jobs, mostly in health care), along with massive downward revisions of the May and June numbers by a combined 258,000 jobs.

Revisions like this are standard practice, and usually, they make me feel more confident in the BLS process. It’s a regular admission of the obvious: no data collection method is perfect. The BLS releases the best information they have in a timely way, and then they update their data as more accurate numbers become available.

This particular revision finally aligned the data with what TA professionals have been experiencing on the ground. I’ve long trusted the integrity of BLS data, even when conspiracy theories spread during the last two administrations that they were cooking the books for political reasons.

That trust took a major hit when, on the same day that the revised data was published, the Trump administration fired BLS Commissioner Erika McEntarfer. If this administration shoots the messenger when the message is inconvenient, how can we trust that the next messenger will continue to deliver the truth?

Which AI model is in your TA Tech Stack?  

According to a new study from Gartner, just 26% of candidates trust that AI can evaluate them fairly, although I wonder how those candidates would rate recruiters at doing the same job.

Those candidates’ fears about AI competence are well-founded, according to new research by Kevin Webster. Kevin, a data analyst, tested a number of leading AI models to see if they could tell the difference between a Director of Finance resume and a waiter’s resume stuffed with “finance-sounding” keywords. Some models, like ChatGPT, did a good job. Others failed the test: Grok rated the keyword-stuffed candidate a 92 out of 100.

These tools are making real decisions that impact people’s lives. As adoption accelerates, you have to ask:

Do you trust the model inside your TA tech stack? Do you even know which one it is?

As for me, I’ve placed my own trust in AI (ChatGPT) to plan my vacation. So far it’s done a pretty good job, except for a few minor (but irritating) things like recommending restaurants for lunch that aren’t open until dinnertime.

I’m skipping next week’s newsletter to enjoy the rest of my time in Spain. Until then: What AI tools are in your recruiting stack, and do you trust them?

David

P.S. If you found this newsletter valuable, chances are your colleagues will too. Feel free to forward it along—and if it landed in your inbox by way of a friend, you can subscribe here to get the next one directly.

Featured Story

AI Gave a Waiter a 92% Score for a Finance Director Job. You Should Be Worried.

I mentioned this article above in the newsletter, but I’m highlighting it here too because we see so many opinions about AI competency, and so few people actually taking the time to do the testing. Kevin did, and he brought receipts. (ERE)

More Recruiting Insights

Human First: Randstad Abandons Monster Employees with a 49% Shrug. Speaking of trust… If I was a Randstad employee right now, I’d be wondering why I would not be treated this way on the way out. (Job Board Doctor)

Microsoft study shows jobs most — and least — impacted by generative AI. I’ll admit to being a bit relieved to not see HR or recruiting on this list. It’s cold comfort though, since all knowledge work is under pressure. Physical labor is back, baby! (GeekWire)

US woman jailed over North Korea remote-work scam. A few weeks ago, I ran a brief about the remote-work scam where North Korean operatives would get remote work at US companies and then use those positions to conduct corporate espionage and funnel money the the North Korean regime. This article tells the story of Christina Chapman, a US Citizen now serving eight years for her role in the scam. (The Guardian)

Hiring Comes to a Halt in July. A solid summary of the July BLS data by SHRM’s Roy Maurer. As he points out, outside of the pandemic year in 2020, this year has been the weakest for job creation since the 2007-2009 financial crisis. I expect that further revisions are going to make those numbers even worse. (SHRM)

Judge orders Workday to supply an exhaustive list of employers that enabled AI hiring tech. The latest in the ongoing Mobley v. Workday case is that Workday must must propose a timeline for identifying all customers who used HiredScore to score, sort, rank, or screen applicants by August 20. (HR Dive)

SAP to acquire SmartRecruiters to enhance its SuccessFactors HCM suite. The big news in recruiting tech this week was SAP’s intent (pending regulatory approval) to acquire SmartRecruiters. SmartRecruiters, last valued at $1.5 billion, brings with it over 4,000 customers and a solid reputation for it’s technology and UX. (CIO)

Conferences

ERE Recruiting Innovation Summit

San Diego, CA

November 4-5, 2025

The ERE Recruiting Innovation Summit is the premier independent, practitioner-led event designed to deliver actionable insights for talent acquisition professionals. Join your peers to explore practical ideas, proven best practices, and case study solutions from leaders who are tackling the same recruiting challenges as you.

The full agenda is now live—see what’s in store and start planning your experience.

Register now and save $200 with early bird pricing, available through August 31.

Webinars

Scaling Smarter: Maximizing Talent Acquisition Efficiency with Limited Resources

August 14, 2025 | 2:00 PM EDT | 1 Hour

Learn how top talent acquisition leaders can scale hiring efforts without burning out their teams or going over budget. We’ll discuss practical strategies companies are using to fill roles faster, elevate candidate experience, improve candidate quality, and simplify hiring workflows — all while keeping the experience human. You’ll hear how modern TA teams are using smart tools, streamlined processes, and focused engagement to reduce manual work and quickly connect with the right candidates. (ERE)

The Talent Magnet: Cultivating a Company Culture that Attracts and Retains the Best People

August 20, 2025 | 2:00 PM EDT | 1 Hour

The best job descriptions can’t compensate for misaligned values. Even the most competitive benefits fall short when culture is inconsistent. This webinar will explore how building a strong, intentional culture can help organizations attract and retain top talent. We’ll cover practical strategies for aligning culture with hiring goals, strengthening employee engagement, and creating a workplace that draws in the best people. (ERE)

Tech Demo Tuesday: Discover the Hottest Tools in Recruiting

August 26, 2025 | 2:00 PM EDT | 1 Hour

Join us for Tech Demo Tuesday, a dynamic webinar series designed for recruiting leaders and talent acquisition practitioners eager to explore cutting-edge technology in the recruiting space.

This week we’ll have Adam Couch and Ian Nadarajahfrom Oleeo, who will demo their AI-first platform including their ATS, CRM, and chatbot OleeoQ.

We’ll also be joined by Survale CEO Jason Moreau, who will show how their talent feedback platform can gather candidate feedback and use AI analysis to identify actionable insights. (ERE)