AI Picked the Best Resume. It Didn’t Pick the Best Candidate.

Why AI in recruiting is useful, but context, relationships, and human judgment still determine who will succeed

It’s been an interesting time seeing artificial intelligence (AI) make its way into professional recruiting. It saves me time and effort on a daily basis. It helps me do so many things: organize information, improve job descriptions, identify potential skill matches, develop interview questions and summarize notes. I love using it to transcribe interviews so I don’t have to take notes while people are talking – I can focus just on them and what they are saying. AI can make recruiting faster and smarter, but it can’t replace the human insight that leads to a truly great hire.

Recently, I presented three candidates to a longtime client for an interim accounting and finance role. This client is a big proponent of artificial intelligence. As soon as he received the resumes, he uploaded them, along with the job description, to Claude and asked it to identify the best candidate.

Claude selected Candidate No. 2.

Based on the resume alone, I understood why. On paper, the candidate had the right experience, the right technical background, and many of the qualifications listed in the job description.

But I disagreed with Claude’s assessment.

The candidate could do the work. But did he have the right work style and fit for this particular client, team, and situation? Based on my six years of working with the company, I didn’t believe his communication style and approach would work as well as those of the other candidates.

I shared my concern with the client before he interviewed all three candidates.

Afterward, he said, “You were right.”

Claude was asked to identify the best candidate with only the resume to go on. Essentially the question to Claude was: “Which resume best matches this job description?”

The real question, in terms of assessing the “best” candidate, is: “Which person is most likely to succeed in this role, with this team, at this company, at this particular moment?”

AI can answer the first, but not the second (yet!). A skilled recruiter knows which context matters and uncovers information that may never appear in a resume. AI can’t account for this context.

AI had three resumes and a job description. I had six years of context.

AI Only Knows What You Give It

Claude had three resumes and a job description. I had six years of context.

I knew much more about this hiring process than AI would ever know. I understood the personalities and working styles within the accounting and finance group. I knew how the team interacted, where communication had broken down in the past, and what kind of leadership the group responded to best.

I also had broader historical knowledge. I knew the company’s history with its systems: what had gone well, what had gone badly, and which issues still created frustration. I intuitively understood what the client meant when he described the kind of person he was looking for, including the things that would never make it into the formal posting.

That institutional knowledge matters. A job description never tells the whole story, and it isn’t meant to. What you won’t find there are the realities surrounding the role:

• The team is still recovering from a difficult restructuring.
• The CFO wants someone who will challenge assumptions without alienating the staff.
• The role requires a person who is strategic but also willing to work in the details.
• The company says it wants transformation but may not yet be ready for rapid change.

A good recruiter gains the trust of their clients and learns those things over time. An AI system can’t factor them into a recommendation unless someone first recognizes that they matter, and then shares them. But there is a wealth of human experience and intuition that can’t be shared.

AI may eventually become much better at interviewing and assessing work style, motivation, and drive. I might get better at entering more context! But I don’t believe today’s tools can fully replace the judgment that comes from knowing the client, the people, the history, and the operating environment. To make the same call, an AI system would need accurate context about all of those factors and would then need to interpret that context correctly.

As an example, a strong recruiter presenting three candidates should be able to tell you:

• Candidate A is the strongest technically, but may need support adapting to the pace of the organization.
• Candidate B has slightly less direct industry experience, but has led a nearly identical system transition and should connect well with the team.
• Candidate C could stabilize the function quickly, although the company may need to clarify the long-term scope of the role to keep the candidate engaged.

That perspective comes from conversations and human insight, not keyword matching.

This is also why I use the word “fit” carefully when discussing candidates. Fit should never mean hiring people who look, think, or behave exactly like the existing team. It should mean evaluating a wide range of softer job-related factors: communication, work pace, decision-making, leadership style, adaptability, willingness to be hands-on, and comfort with the organization’s environment.

Use AI to make recruiting faster. Use an experienced recruiter to make the hire better.

The Best Approach Is Human Judgment Plus AI

Used well, AI gives recruiters and hiring managers more time to focus on higher-value conversations and decisions. As I mentioned at the start of this post, AI has added a lot of value to my work. I talk to more clients and more candidates now than before I started using AI. I love creating a new project and teaching AI to go look for hard-to-find candidates for unusual roles. BUT I spend a lot of time training it before I ask it to search, and I reach out personally to candidates who look promising with a tailored message (not the one AI recommends!).

At its core, hiring is a human activity. AI should support and inform a recruiter’s judgment, not replace it.

I think the best hiring process combines AI’s productivity and speed with a recruiter’s experience, intuition, relationships, and accountability.

That is what companies should be paying a recruiter for.

AI can help you find candidates. We can help you find the right one.

The best hire isn’t always the person whose resume matches the job description the best. If you’re looking for a recruiting partner who brings experience, context, and human judgment to the process, reach out to one of our recruiters.

Connect with a P4P Recruiter: https://p4pconsult.com/contact/