For years, the written application was one of the main sources of information in hiring. A clear, well-structured CV suggested a candidate who could communicate. A thoughtful cover letter showed interest in the role. Recruiters learned to read these documents for signals beyond the content itself.
That has changed. Many candidates now use AI writing tools to prepare their CVs and applications, and it is easy to see why. These tools help people express their experience clearly, adapt their CV to each role and write in a language that is not their own. But they also mean that a polished application says much less than it used to.
The new situation
In earlier articles, we wrote about how keyword filters reject good candidates and about the promise and pitfalls of AI in screening. The current situation adds a new layer. On one side, companies use automated tools to screen applications. On the other, candidates use automated tools to write them. Sometimes the CV is optimised for exactly the kind of system that will read it.
The result is a growing number of applications that look well matched to the role, use the right keywords and read smoothly, regardless of the candidate’s real experience.
This is not about catching candidates
It is tempting to treat AI-written applications as a problem to detect and punish. We think that is the wrong approach. Using a tool to write a clearer CV is not dishonest, in the same way that asking a friend to review a CV is not dishonest. Many strong engineers are not good at writing about themselves, and these tools help them.
The problem is not the tool. It is that the written application has become a weaker signal, and screening processes that rely heavily on it become less reliable.
What becomes less useful
Some signals that recruiters traditionally relied on now carry less weight:
- Writing quality. A well-written CV no longer says much about how someone communicates.
- Keyword match. A CV that closely mirrors the job description may simply have been tailored by a tool.
- Cover letters. Personalised motivation letters can be produced in seconds.
- Automated scores. If the application was written to match the system, a high score means less.
What becomes more important
Screening now needs to rely more on signals that are harder to produce automatically:
- Concrete detail. Specific projects, decisions and trade-offs that a candidate can explain in depth when asked.
- Verifiable work. Code repositories, contributions, published work or references, where available.
- Conversation. A short early call with a person often reveals more than any written document.
- Practical assessment. Exercises that reflect the real work, discussed with the candidate afterwards, show how they think.
- Consistency. Whether what someone says in conversation matches what is written on paper.
Keep the process fair
As screening shifts towards conversation and assessment, fairness needs attention. Candidates who are less comfortable in interviews, or who are interviewing in a second language, should still have a fair chance. Clear questions, structured interviews and consistent evaluation criteria help reduce bias.
Transparency matters too. Telling candidates how the process works, which stages involve automated tools and what is expected of them is respectful and builds trust. It also reflects the direction of European regulation, which places increasing obligations on companies using AI in hiring decisions.
Adjust the volume problem
Because applications are easier to produce, many roles now receive more of them. Responding by adding more automated filtering can create a cycle in which neither side’s documents are read by a person. A better answer is often to reduce the volume at the source: clearer job descriptions, specific requirements and more targeted searching.
The recruiter’s role is growing
In this environment, experienced recruiters matter more, not less. Understanding the role in depth, knowing which questions reveal real experience, and recognising a strong engineer behind an ordinary CV, or a weak one behind a perfect one, are skills that tools do not replace.
This is central to our Precision Hiring approach at Boomit. We combine careful searching with direct conversations and practical assessment, so that decisions are based on real skills rather than polished text. If you are seeing more applications and finding it harder to tell candidates apart, we would be glad to talk. You can reach us at boomit.pt.