AI-on-AI War Rewrites Hiring: Trust Gap Widens

The modern hiring landscape is undergoing a profound transformation, marked by an escalating “AI-on-AI war” where both job seekers and employers are increasingly leveraging artificial intelligence. This surge in AI adoption has created a widening trust gap, prompting companies to fundamentally rethink their recruitment strategies. Roei Samuel, founder of the networking platform Connectd, recounts instances of candidates providing suspiciously perfect answers during video calls, their eyes betraying the use of unseen prompts.

Generative AI has rapidly permeated the job market, becoming a mainstream tool for job seekers. By late 2024, an estimated 68% of European tech workers were actively using AI to refine their CVs or even complete entire applications. Tools like Sonara, LazyApply, and JobCopilot facilitate the rapid submission of dozens of applications daily. Data from TestGorilla in June indicated that 37% of UK job seekers now use AI for applications, a figure that jumps to 60% among early-career candidates, up from 38% the previous year, according to Bright Network.

Despite the pervasive use of AI by candidates, 85% of employers actively accept AI-assisted applications. However, this acceptance does not equate to apathy. Companies are grappling with how to discern genuine talent amidst a deluge of AI-generated content. A Canva survey revealed that 45% of employees across Europe had used generative AI to build or improve their resumes, often with positive results. Yet, trust remains fragile; 63% of UK hiring managers believe candidates should disclose AI involvement in their application materials.

Employer attitudes toward AI in applications vary, often depending on the role. A global survey by Experis found that 28% of tech leaders were comfortable with AI for personalizing resumes or cover letters, 26% for problem-solving tests, and 24% even for answering interview questions. Only 15% deemed AI use unacceptable across the entire application process. Duco van Lanschot, co-founder of fintech startup Duna, exemplifies this nuanced view: while an engineer using ChatGPT to polish a written application might be acceptable, a growth or sales hire doing so would be a “big red flag” given the inherent need for human communication in such roles. Marija Marcenko, Head of Global Talent Acquisition at Semrush, emphasizes that AI hasn’t “broken hiring” but has undeniably reshaped candidate engagement.

In response to this shift, traditional application materials are losing their relevance. Khyati Sundaram, CEO of ethical AI hiring expert Applied, describes CVs as a “broken artefact,” following cover letters into obsolescence. She argues that relying on keyword scanners or AI tools for resume analysis is ineffective because candidates often fail to perform in subsequent interviews. Consequently, employers are increasingly favoring structured questionnaires and skill-based tasks designed to evaluate a candidate’s thinking process rather than just their ability to craft an AI prompt. TestGorilla reports that 77% of UK employers now use skills tests, finding them more predictive of job success than CVs. This skills-based approach, according to LinkedIn’s Economic Graph Institute, could globally expand talent pools by 6.1 times, fostering greater diversity.

Companies are adapting their assessment methods. Semrush, for instance, trains hiring managers to detect AI-generated “fluency without depth” in real-time coding challenges or task-based interviews. They’ve replaced generic prompts with in-depth interviews that delve into experience, soft skills, and thinking patterns, which are “hard to fake, with or without AI.” Applied’s system employs human reviewers for pattern matching, comparing submissions to known AI outputs, rather than relying on often inaccurate AI detectors. Travel scaleup WeRoad actively encourages AI use but balances it with traditional in-person interviews and real-time scenario-based exercises to assess collaboration and cultural fit.

The human element is making a significant comeback in the hiring process. References, particularly “backchannel” conversations with direct colleagues, are regaining importance. Santiago Nestares, co-founder of accounting startup DualEntry, emphasizes the value of face-to-face interactions and deeper reference checks to ascertain how candidates handle pressure and work within a team. Roei Samuel of Connectd notes that candidates are increasingly building social proof to bridge the trust gap.

The much-maligned take-home task is also on its way out. Once a common assessment, these unpaid and time-consuming assignments are now viewed with skepticism by employers due to the ease with which AI can generate responses. Live interviews, technical walk-throughs, scenario-based challenges, and even roleplay simulations are becoming the new standard, particularly in product, design, and marketing roles. Andreas Bundi, founder of Berlin-based HR consultancy Bundls, observes that many companies are questioning the utility of take-homes when live assessments offer a more authentic evaluation. He notes that AI-first companies tend to be more relaxed about candidates using AI tools, though this is rarely explicitly stated. A recent example involved a data scientist who failed an interview by manually wrangling data instead of automating it, missing the company’s preference for strategic thinking over manual execution.

Despite the widespread adoption of AI-powered applications, many companies still lack formal guidelines for AI usage in their recruitment processes. BrightNetwork reports that 40% of employers using their services haven’t set such guidelines, though 28% plan to for the next recruitment season. Of those with guidelines, 44% disallow AI use by candidates. Sundaram notes that “vanguard employers” are beginning to assess AI literacy, even adding questions like, “How will you use AI in this job?” However, she cautions against ethically questionable AI tools, such as video screening with facial tracking or voice sentiment analysis, which raise significant privacy concerns.

Instead, Sundaram argues, the solution lies in redefining what candidates are being tested for. Applied has shifted from traditional “job architecture” to “task architecture,” evaluating not only skills but also essential human traits like resilience, adaptability, and mission alignment. These qualities, she contends, will become even more critical as jobs evolve, especially in startups where versatility is key.

Generative AI is fundamentally reshaping the hiring process. Forward-looking startups are not resisting this change but are building more effective processes around it. While CVs may be outdated and applications increasingly synthetic, the true differentiator remains inherently human: the capacity to adapt. Companies that understand this and structure their hiring accordingly are not just future-proofing their teams; they are actively rewriting the rules of work for the AI era.