A striking contradiction is emerging in the job market for AI talent. AI has made applying almost free, and application volume is up sharply, while companies struggle to find the right AI talent.
On the one hand, interest in AI has never been higher. More people are learning about the field, building relevant skills, and applying for AI-related roles. On the other hand, as the market matures, the threshold for those roles is rising, as companies increasingly expect candidates to combine strong technical foundations with practical experience, business understanding, and the ability to deliver measurable results. This creates a paradox: organizations receive more applications than ever, yet many still struggle to identify the truly strong AI talent they need.
AI-written applications make screening harder
AI has collapsed the cost of applying. Candidates can now generate polished CVs, tailor them to individual vacancies, and optimize their wording within minutes using CV builders, general-purpose assistants such as ChatGPT, and the tailoring tools built into job platforms. On paper, these applications can look highly relevant. However, when recruiters compare them with a candidate’s LinkedIn profile, experience, or interview performance, the match is often much weaker.
The term “CV shooting” was often used in a recruitment agency setup: sending one broadly written CV at every open role and letting volume do the work. Applicants now do it themselves, at a scale no agency ever managed — and increasingly they delegate it to an agent that applies on their behalf. The result is a higher volume of applications, a lower-quality inflow, and a much heavier screening burden.
What happens when both sides automate?
This makes recruitment more demanding and time-consuming. The obvious response is for recruiters to use AI as well: to screen applications, compare profiles, identify inconsistencies, and prioritize candidates. While this can improve efficiency, it introduces another tension. Candidates use AI to increase their chances of being noticed, while employers use AI to filter them out. At the same time, automated rejection processes frustrate many applicants, who see decisions that feel impersonal and that miss their potential. Both sides risk entering an arms race in which more automation creates more volume, more filtering, and less meaningful interaction. This leaves both parties frustrated with each other's AI use.
Extra note: There is a legal floor linked to this as well. AI systems used to screen, rank, or select candidates are classified as high-risk under the EU AI Act, which brings obligations on risk management, bias testing, logging, transparency, and effective human oversight. And do not forget about GDPR, which already restricted solely automated decisions with a significant effect on individuals. Taken together, the two regimes make human review of every rejection a requirement rather than a nicety—which happens to be the same conclusion good hiring practice reaches on its own. But these are guidelines for companies; what about individuals using AI to apply?
How to use AI without losing authenticity
The way forward is not to reject AI, but to use it without losing authenticity. Candidates should use AI to communicate their real experience more clearly, not to present themselves as a match for roles they are not suited for (we avoid going into the rabbit hole of misalignment on it in this blog post). Recruiters should use AI to reduce administrative work, not to remove judgment, curiosity, and personal contact from the process. Companies can help by writing clearer vacancies, defining which requirements are essential, and creating application processes that reward evidence of capability rather than keyword optimization. In a market shaped by AI, trust and human connection are the scarce assets.
How ML6 changed its own hiring process
ML6 is testing a few trials to counter the shift toward bulk and agent-driven applications. Rather than relying only on a CV, we ask candidates to answer a few additional questions that are difficult to complete meaningfully without genuine reflection. This can include a short video introduction or a concrete example of a project they are proud of, including their personal contribution and what they learned. These steps give candidates room to show the person behind the application. They also help us distinguish between a polished, AI-optimised profile and someone with a sincere interest in the role who can speak authentically about their experience. Comparing Q2 2026 to Q1, we saw a 45% increase in applications, with only a 12% increase in candidates reaching the offer stage. Since ML6 doesn't use AI to screen incoming applications, this really shows the burden AI-assisted applications place on recruiters.
The organizations and candidates that combine technology with honesty, relevance, and genuine dialogue are most likely to find the right match. See how we hire and what we look for on the ML6 careers page.
