Job Search: AI Allowed, Bots Risky

Recruiters and coaches describe a fine line: AI can help structure a resume and prepare for interviews, but it cannot industrialize interchangeable applications. Greenhouse observes that in the third quarter of 2025, an average job posting received nearly three times more responses than in 2017. In this context, attempting to bypass the ATS or automate submissions undermines credibility, especially when verifiable achievements make the difference. An example is provided with Lamar Rogers, who equips himself without delegating the application itself.

More Applications per Job, but Bots Remain a Bad Bet

In the third quarter of 2025, Greenhouse found that a typical job posting received almost three times more applications than in 2017, at a comparable unemployment rate. Candidates are therefore submitting more applications, intensifying competition for each position.

However, using a bot to apply en masse is not the most effective way to secure an interview. Stephen Telford warns that if automation is apparent, it may suggest that the position was not worth the candidate's effort. In a tight job market, generic applications have little chance of success.

Don’t Hide AI, but Prove Every Claimed Result

Coaches believe there is no need to conceal the use of AI, as long as the information remains accurate. For them, employers prioritize the truthfulness of achievements over the formal perfection of the resume. Stephen Telford summarizes the requirement: it doesn’t matter if AI helped with the writing; the key is that the claim holds up under scrutiny.

He cites the case of a 30% increase in sales: if the candidate cannot detail the "how" or substantiate this growth (for example, on LinkedIn), their credibility erodes. His advice: before allowing a tool to rephrase an achievement, ensure that one can explain the mechanism, the starting point, the actions taken, and the result. Telford, who has previously recruited for AWS, recalls this framework, while providers like Anthropic claim to embed an imperceptible watermark in the text generated by their Claude models. The use of tools is thus becoming more detectable, reinforcing the importance of verifiable claims.

Demonstrable Authenticity: Personal Brand and Consistent Language

In the face of cover letters generated by tools and filled with the same buzzwords, maintaining a personal brand helps to stand out. Liz Andora (DHI, parent company of Dice) sees AI as a support for structuring a resume or polishing language, while emphasizing that the application must remain true to the individual.

Recruiters look for signs of reality: up-to-date LinkedIn profiles and a clear record of accomplishments. Ron Porter (Korn Ferry) says he receives impeccably formatted documents that do not resemble their authors. The language must reflect the candidate's communication style and remain something they could have written themselves.

Bypassing the ATS Can Be Detrimental: Prioritize Consistency and Internal Support

Some attempt to deceive ATS filters by hiding keywords in white text. Liz Andora acknowledges that passing these systems is a real challenge, but warns: injecting keywords that do not match the skills can backfire on the candidate. Stephen Telford sees this as a signal of a quest for visibility at the expense of credibility; if the list does not align with LinkedIn, GitHub, or the provided explanation, he doubts the entire application.

As employers already filter through AI and candidates use it for writing, Xenia Wade finds the frustrations understandable: companies fear texts that are too "AI" while filtering with algorithms. One way to break free from this is to obtain an internal recommendation; Andora advises relying on a contact within the company to ensure the application is reviewed.

How Far to Delegate to AI: The Choice of Lamar Rogers

Laid off in March, Lamar Rogers says he has equipped himself extensively: adapting his resume with Claude, an AI-driven dashboard for each interview, and even developing a proofreading tool. He claims nearly 200 applications supported by these techniques, but refuses to let a machine apply on his behalf. For the senior positions he is targeting in finance and technology, entrusting the entire process to an agent seems unacceptable to him.

At 34 years old, he explains that he spent months iterating until he achieved a resume, cover letter, and process that met his expectations. His principle remains the same: AI does not do everything; he controls the content and retains the final decision. For him, the tool provides structure, but the initiative and management of the application must remain human.