Growth: Job Searches in the Age of AI

Looking for work has always required a strange mix of self-promotion, endurance, and timing. Now, job seekers have another element to contend with, and it moves faster, sees differently, and operates largely behind the scenes: artificial intelligence, or AI.

AI is involved in more hiring decisions than most job seekers realize. Beyond scoring applications and resumes using pattern recognition and user-defined algorithms, AI is now analyzing and even conducting video interviews and nudging certain candidates forward while dismissing others before they're ever seen by a human.

For job-seekers, AI can help applicants tailor documents, write emails, and search more efficiently, but it can also create challenges when employers use AI tools to detect AI-generated content and disqualify those candidates.

That duality is part of the problem. Companies rely on automation to move applicants through the funnel, and then flag resumes that appear too polished or too similar to others produced by the same tools. Employers expect personalization but often don't offer the same in return. Worst of all, the rules of engagement are rarely made public.

Current job seekers find themselves trying to navigate a system that is increasingly opaque. Let’s examine how AI is changing the job search process at every stage, from application to interview, and how to use these tools without being mistaken for one.

AI Resume Screening

For most large employers, the first resume and application read is done by a machine. Systems, known as applicant tracking systems (ATS), are designed to process high volumes of applications efficiently. At the most basic level, they search for keywords that match the job description. More advanced ATS platforms extract data from resumes, interpret that information using natural language processing, and rank or sort candidates based on internal logic or external criteria. Some even attempt to infer personality traits, decision-making patterns, or career readiness from how applicants describe their previous roles.

These systems, often marketed as tools that reduce bias, are frequently trained on historical hiring data that reflects decades of inequality. If an algorithm learns to identify top performers by analyzing the resumes of a company's current employees, and those employees were mostly hired through informal networks or represent only a narrow demographic slice, then the system will learn to prefer similar applicants. It will flag those who differ simply because they do not fit the encoded pattern.

In this eye-opening episode of BBC Newshour, “How AI is Changing the Job Marketplace,” Investigative reporter Hille Shellman shared an instance of an employer who fed AI the resumes of people already employed by the firm, with instructions to find common patterns in those resumes, and use those metrics to filter incoming resumes so that only those that were similar would pass through. Shellman went on to explain: 

[The resume screening tool] had “learned” that the name Thomas was predictive of success at the company. So, [applicants with] the word Thomas on [their] resume, got more points.(…)Obviously it's not meaningful, right? It's just arbitrary.

Just as concerning as the use of AI to highlight “important” words, is the use of AI to downgrade words. Downgrades can include overused words or cliché phrases, and even, in a highly a discriminatory practice, can include words that are gendered or believed to have racial overtones. 

For example, in the same BBC interview referenced above, participants discussed examples of resumes that were downgraded if the word “women” appeared, and, in one case, reported that “If [an applicant] had the word baseball on [their] resume, [they] got more points. If [they] had the word softball on there, [they] got fewer points. Again, pointing to gender discrimination.” 

More benign, but still frustrating, ATS can also be tripped-up by various types of formatting like headings, fonts, or file types. Dense blocks of text might confuse the parser, and custom graphics or multi-column layouts are often unreadable. A resume that would catch the attention of a human hiring manager might not make it past the AI gatekeeper.

Crafting Cover Letters

The modern job seeker is expected to write with clarity, confidence, and a degree of personality that demonstrates both professionalism and cultural fit. They are also expected to do so repeatedly, often for dozens of roles that differ only slightly from one another in title or focus. This is where generative AI tools have really become popular with job seekers. Even the most talented, driven, and highly motivated candidate can become overwhelmed at the sheer volume and pace of applications required.

Programs like ChatGPT, Claude, and Copy.ai are now widely used to draft cover letters, suggest phrasing, or reword bullet points in ways that feel polished and professional. These tools excel at generating grammatically sound text that sounds appropriately enthusiastic and echoes common business language that aligns with corporate tone. 

However, AI's reliance on hollow, broadly applicable sentiments, often padded with filler phrases and the use of repetitive structures in a weak attempt to "sound human," is a dead giveaway to their use. Many companies have become wary of job seekers who rely too heavily on AI tools, and have begun using separate software to screen for what they consider "AI-written" resumes and cover letters.

Using AI to Stop AI 

Ironically, some of the screening tools used to detect AI text, are powered by the very models used to generate the cover letters and resumes in question.

These detectors look for patterns in syntax, phrasing, and structure that resemble large language model outputs. If a resume or cover letter scores too high in sentence uniformity, semantic repetition, or other hallmarks of machine-generated content, they may be filtered out, no matter how relevant the experience. AI-driven decisions happen without explanation. Applicants are left wondering if they were underqualified, overlooked, or simply flagged by a tool they didn’t even know was being used.

There is a smarter way for job seekers to use these tools. Rather than outsourcing the entire task, applicants can use AI to suggest structure, identify stronger verbs, or help reframe a paragraph that feels flat. The key is to build on that scaffold with details no machine could fabricate. Consider including specific references to the company's mission, direct connections to the role's requirements, or insight drawn from recent news about the organization. 

AI Interviews

Some companies have begun using AI-powered platforms to conduct and evaluate first-round interviews. These aren't Zoom calls with a recruiter (which can also be tedious), but actual, pre-recorded video interviews, where candidates answer questions on camera without anyone on the other side. The footage is then analyzed by yet another machine, using software trained to detect signals that hiring managers rarely name and that candidates are never taught.

The system watches more than listens. It tracks the timing of each response, measures facial expressions, categorizes vocal tone, and even monitors blinking patterns. Some platforms attempt to gauge confidence, enthusiasm, or claim to detect emotional intelligence. These systems are actually measuring proximity to a trained model. Just like in resume filtering, they are programmed to look for traits that resemble past successful hires or that align with the employer's performance metrics, creating a hidden set of expectations that can seem contradictory. 

For example, speaking too quickly can be penalized, but long pauses may register as uncertainty. Speech patterns that deviate from the norm, whether due to regional accent, cultural rhythm, or neurodivergence, might be interpreted as a lack of clarity or a poor fit. If the system is scanning for energy or composure, candidates who are naturally subdued or anxious on camera may be marked down — even if their skills are exceptional. Even lighting, camera angle, and background can influence perception.

This kind of invisible exclusion is especially harmful for candidates from nontraditional backgrounds, including career changers, older workers, or those whose education and experience fall outside the typical corporate mold. It also presents challenges for individuals whose communication styles diverge from what AI has been trained to view as "confident" or "engaging." This group often includes neurodivergent professionals, international applicants, and those who speak multiple languages.

There is no guaranteed way to perform well under these conditions, but preparation can help. Using a structured framework, like the STAR method (Situation, Task, Action, Result), can keep responses focused and coherent, reducing verbal filler and trailing thoughts. 

Ultimately, the best strategy is to prepare as if speaking to a skeptical algorithm, accept that this stage is a data capture session, not a true conversation, and the person reviewing it may not be a person at all.

Tips for AI Bias and Gatekeeping

The promise of automation in hiring is built on the idea that machines can make faster, more objective decisions than people. But, far from being neutral arbiters, algorithms are created to reflect the values and assumptions of the people who build them. Meaning that, in the end, they can (and do) replicate the same patterns of exclusion that human decision-makers have long struggled to overcome.

When a company uses an AI-driven system to evaluate cover letters, resumes, or video interviews, the criteria for advancement are rarely visible. Applicants are not warned when an AI interview platform flags their speech cadence or vocal tone for failing to match a trained model's expectations. They are not informed when their resume is ranked lower because it deviates from the language used by previous hires, and they certainly aren't told that their correspondence contains words that are being discriminated against. 

To counter these unfair practices and clear AI-related hurdles, job seekers need to employ strategic methods. Listed here are a few tips that can help job seekers use AI to their advantage, while bypassing the screening tools and dodging some of the ATS metrics.

For AI-enhanced resumes and cover letters:

• Use standard formatting, avoid text boxes or graphics.
• Mirror the language used in the job description, but don’t copy sentences outright. 
• Resist the urge to embellish. 
• Avoid over-optimization. If a sentence feels too perfect, it probably is. Let a few “rough edges” show your personality or thought process.
• Embed details that can’t be fabricated: specific projects, unique metrics, individual achievements. 
• Breaking up overly balanced sentence structures, insert language and add rhythm and pacing that reflects human thought and speech patterns.
• Clearly label sections with simple headings like "Experience" or "Education." 
• Rephrase AI-generated content to sound like you, and vary the length of your sentences.
•  Avoid “negative keywords.”

And for AI interviews:

• Treat asynchronous interviews as structured performances.
• Practice answers aloud and record yourself to catch pacing or tone issues.
• Use tools like Google’s Interview Warmup to simulate questions, then edit your content and delivery. 
• Maintain eye-level camera framing, natural lighting, and neutral backgrounds. 
• Keep answers concise; tangents or rambling can be interpreted as disorganization.

Finally, be sure to track your application outcomes to note repeated patterns of rejection despite qualification. 

Applicants should also focus their energy where algorithms cannot compete. Following up with a brief message on LinkedIn, introducing yourself at a virtual networking event, or joining an industry-specific forum to create a line of connection that no resume can replicate. Small moments of interaction can determine which names rise above the database and into consideration.

Remember. . .

AI is changing and shaping the hiring process, but people are still making the decisions that matter. The strongest candidates aren't the ones who completely avoid technology or lean on it entirely. They're the ones who learn how to use it with discretion, edit with intention, and show up as unmistakably real, unique, and irreplaceable.