Evidence vs claims: why assessing gains relevance in the age of AI

In the age of AI, claims have become cheap and evidence has become valuable. When anyone can generate a flawless CV, cover letter, or answer, what’s claimed no longer differentiates candidates. What’s assessed does: evidence is observed under the same criteria and once again sets one person apart from another. That’s why assessing before interviewing gains relevance, not loses it.

For years, hiring rested on claims: the CV, the cover letter, the account of accomplishments. They worked as a filter because writing well took effort, and that effort was itself a signal. AI changed that economy: today a polished claim is cheap and abundant. And when something is cheap and abundant, it stops discriminating.

The economics of the signal flipped

It’s useful to think about the change in terms of supply. Before, well-crafted claims were scarce, so they served to distinguish. Assessed evidence existed, but many processes saw it as an extra step. Today the relationship is reversed:

• Claims got cheaper. Generating a correct text no longer demonstrates effort or any particular ability.
• Evidence rose in value. Precisely because it can’t be “written,” it remains scarce and, therefore, useful for differentiating.

What counts as a claim and what counts as evidence

	Claim	Assessed evidence
Examples	CV, cover letter, self-description	Result of a test for the role
Origin	What the person says about themselves	What’s observed when they respond
Cost to produce today	Low (via AI)	Stable: requires taking the assessment
Comparability	Low	High, with common criteria

The right-hand column is the one that still gives you filtering power once the left-hand one has become uniform.

Why assessing before interviewing gains relevance

Assessing early does three things the CV no longer can:

• It restores filtering power. It separates candidates on the same scale, not on different narratives.
• It gives context to the interview. You reach the conversation knowing what to look at, instead of discovering it live.
• It makes the decision defensible. Comparing with common criteria leaves a trail of evidence, not of impressions.

None of this implies that evidence predicts the future. It describes styles and abilities in a comparable way; it doesn’t guarantee results. You can see how this rigor holds up in the science section or explore the available tests in the library.

AI interprets, the person decides

AI isn’t the villain of this story, nor the judge. On the candidate’s side, it helps with writing; on the recruiter’s side, Kokoro’s AI helps interpret and organize the results according to the role. In no case does it decide. The final decision belongs to the human team, and assessed evidence is what supports it.

In summary

AI made words cheap and, with that, stripped claims of their filtering value. Assessed evidence kept that value because it can’t be written: it’s observed and compared. Assessing before interviewing isn’t an extra step in the age of AI, it’s the step that recovers the signal. Explore the product or the assessment library.