AI Can Help You Find the Problem. It Cannot Make You Do the Work.

I recently tried to help an experienced professional who had been looking for work for a significant period. Their background included telecommunications, infrastructure, federal IT, and project management. At first, the obvious question seemed to be: What is wrong with the résumé?

Instead of endlessly rewriting résumés, I started examining the job search as a process. Approximately 75 applications had produced 12 recruiter interviews, two hiring-manager interviews, and no offers. These figures did not prove why the search was stalling, but they suggested that the résumé was getting some attention and that the next useful inquiry was what happened between recruiter and hiring-manager conversations.

That changed the question from how to beat the ATS to where the process was losing momentum.

An applicant tracking system, or ATS, helps employers receive, organize, search, and manage applications. Some systems offer parsing, matching, or automated screening capabilities, but their functions and employers’ practices vary. A human resources information system, or HRIS, is broader: it may manage employee records, benefits, payroll, performance, and recruiting integrations.

Neither term means that every application is decided by a mysterious AI gatekeeper. Hiring decisions can involve technology-assisted screening, recruiters, hiring managers, interviews, and organizational judgment. Understanding those distinctions matters more than chasing viral tricks to defeat an algorithm.

AI can compare a résumé to a job description, identify relevant language, distinguish responsibilities from accomplishments, prepare interview questions, suggest alternate roles, and help analyze where a candidate advances or stalls.

But it should not invent achievements or inflate budget authority. There is a meaningful difference between managing a $100 million program and coordinating delivery activities within one. A dollar figure may describe revenue, an account, a portfolio, or a budget. Accuracy matters.

The next step was not another résumé. It was a deeper conversation about what the candidate had actually accomplished, how those achievements could be verified, and how the candidate described leadership, risk management, and difficult stakeholder situations in interviews.

That assessment required the candidate’s participation. The candidate chose not to continue. I cannot know whether confidence, technical gaps, interview technique, market conditions, or some other factor explained the results, because we never completed the assessment.

But the experience clarified a principle: AI can help identify the work. It cannot make the human do the work.

A lawyer still has to exercise legal judgment. A business owner still has to make decisions. A real estate professional still has to understand the transaction. A leader still has to lead.

AI can create a polished output quickly, but the output is not proof of expertise. It cannot verify an accomplishment a person will not explain or substitute for honest self-assessment. Governance before automation means keeping facts, decisions, and accountability with the people responsible for them.

I often say: Experts Drive AI. Sometimes the most valuable result of using AI is not a better document. It is a better question.

In this case, the question began as: How do we fix the résumé? The better question became: What does the evidence tell us about where this process is breaking down?

Use the technology. Study the evidence. Improve the process. Automate what should be automated. But when human judgment, honesty, accountability, or action is required, recognize that the next step belongs to the human.

AI assists expertise. It does not replace participation.