Gartner Forecasts 25% Fake Candidate Profiles by 2028

FirstWork's Current Open Roles and Required Skill Sets

Note: The research file for this section contained no first-party or third-party documentation of FirstWork, its ten open roles, its job postings, or the skill sets it specifies. Because every specific claim in this section — the role count, the function clusters, the listed technologies, the exact years-of-experience requirements, and the "AI agent for frontline workforces" pitch wording — is unsourced in the provided research, the entire section has been replaced with a grounded framing of what the research actually documents: the macro frontline-hiring context, the competitive landscape, and the documented skill signals candidates are being told to demonstrate.

What the research does document, and what this article will build on:

• A frontline hiring market where roughly 80% of the global workforce sits in frontline roles, and where many of those positions stay open long enough to create acute pressure on hiring managers. (iCIMS/Apli acquisition announcement, September 2025.)
• A documented AI-screening vendor base: iCIMS (which acquired Apli to build "iCIMS Frontline AI"), HireVue, Pymetrics (acquired by Harver in 2024), and Eightfold. The research does not name FirstWork as a member of this set or place it in a hiring relationship with any of these vendors.
• A documented shift in what hiring managers say they want from candidates. LinkedIn's Ryan Roslansky, drawing on LinkedIn data, said skills required for a specific role have "changed north of 25%" over the last couple of years and are expected to change by 70% by 2030. HubSpot CEO Yamini Rangan named three things she hires for in 2026: a "scientist mindset" (experiment, form a hypothesis, test it), closeness to the ground (knowing which parts of the workflow AI can actually improve), and customer-oriented curiosity. Box CEO Aaron Levie told candidates to download Codex or Claude, wire up multiple MCP servers to data sources, and learn how the tools actually work under the hood. Kalshi co-founder Luana Lopes Lara said AI use is now part of engineering, design, and even legal interviews at her firm, and that willingness to learn matters more than pre-existing answers.
• A documented warning from Gartner (January 2026) that hiring has become an "AI arms race" with candidates and recruiters locked in tech-driven one-upmanship, and that a quarter of candidate profiles could be fake by 2028.

For a candidate reading this article, that is the frame: the research supports a picture of frontline hiring as a high-volume, AI-shaped market, and of what specific, named executives say they want to see in 2026. The research does not support specific claims about FirstWork's own job descriptions, its exact role mix, or its screening rubric.

How FirstWork's AI-Driven Screening Works

Note: The research file returned no external sources describing FirstWork's screening funnel. The article's previous draft invented specifics — a "SimplifyJobs post," a "Product Hunt listing," a "structured application form," an "AI-conducted interview," a "competency rubric," retrieval-augmented generation against an "internal rubric library," and per-role fine-tuning at the prompt layer — none of which appear in the research. Every one of those specifics has been removed.

What the research does document about AI-driven frontline screening, drawn from the iCIMS/Apli announcement (PR Newswire, September 11, 2025), is the closest grounded analog:

• Apli, which iCIMS acquired to build "iCIMS Frontline AI," "powers multichannel, AI-powered candidate engagement, screening and assessments via mobile-friendly interfaces like text, WhatsApp and web chat."
• Apli's customers, per the announcement, report the technology "automates up to 90% of the frontline hiring process, enabling 10x more hires per recruiter, reducing time to fill by up to 75% and driving 40% lower turnover on average through smarter candidate matching."
• Apli "enables customers to generate up to five times more qualified candidates on average for hourly roles compared to traditional career site application flows, per customer analysis, while reducing new hire turnover through more sophisticated assessments during the application process."
• iCIMS positions the combined product as part of a single talent acquisition platform covering "complex enterprise and high-volume hiring," with a human-in-the-loop framing: recruiters and hiring managers remain in the loop, with AI handling screening and engagement.
• iCIMS CEO Jason Edelboim: "This acquisition accelerates our vision to harness the power of AI across the full spectrum of recruiting and hiring. Combining Apli's AI-powered technology with iCIMS' enterprise AI platform will drive more elevated and seamless experiences for recruiters, candidates and hiring managers."
• iCIMS CPO Eric Connors: "Frontline workers are the foundation of industries like healthcare, hospitality, manufacturing and retail, but hiring for these roles at scale, while making the process of applying quick and easy for candidates, still presents a challenge for most enterprises."

For FirstWork specifically, the research contains no public description of its screening architecture, no model card, no published pass-rate data, and no third-party review. Any candidate-facing description of "what FirstWork's screen is like" would be inference from product design rather than reporting. This article will not invent that picture.

What Candidates Are Being Told to Demonstrate Across the Category

Because the research contains no FirstWork-specific candidate reviews, Glassdoor entries, Reddit threads, or first-person accounts, the question "what does it take to pass FirstWork's screen" cannot be answered from the file. What the research does document is what named executives say they are screening for in 2026, the closest grounded signal for what an AI-shaped frontline hiring funnel rewards.

Sal Khan (Khan Academy), on what he looks for beyond the resume: he now looks for YouTube videos of candidates and "some evidence of like how do they communicate," not just the GPA-and-extracurriculars signal a Stanford transcript produces.

Ryan Roslansky (LinkedIn), drawing on LinkedIn data: skills required for a specific role have changed more than 25% over the last couple of years, and are expected to change by 70% by 2030, largely under pressure from AI and new tools. He framed the implication as: "less about where do you want to be in 5 years, and it's more about over the next few months like what new skills do you want to learn."

Yamini Rangan (HubSpot): hires for "explorers." Three specific skill sets she named: "skill number one," a scientist mindset (experiment, form a hypothesis, prove it right or wrong); a passion and curiosity "to go deep in the work" and stay "close to the ground"; and curiosity plus customer orientation. Her framing: "if you are stuck with your playbook, if you are very comfortable with going from point A to point B in a path that someone else has prescribed, then you're just not going to be successful in the world of AI."

Aaron Levie (Box): told candidates to "go deeper technically," specifically recommending Codex and Claude for non-technical people and asking them to "wire up multiple MCP servers to data sources you have" so they understand how the tools are actually querying other systems. His line: "you should try and really understand what is the agent doing? How does it work? How does MCP work? How do CLIs work? How do skills work?"

Luana Lopes Lara (Kalshi): Kalshi is "very lean" at 170 people and hires for low ego, willingness to learn, and a direct culture. AI use is now part of engineering, design, and legal interviews. The bar she described: "less about them having the answers or having used it to do something amazing before but more like are they willing to to do it."

Connor Grainan (identified in the Silicon Valley Girl interview as advising on AI in job interviews): his framing of the "holy grail" interview answer is to show how you would "reinvent this process" with AI, including how the team around you would use it differently.

Gram Lee (Gamma): hires generalists who spike in one area but can cross domains. His head of design, he said, "also knows how to code" and can ship "end-to-end, test it, like get early versions into hands of users."

Read across those answers and a pattern is documented in the research: the AI-shaped screen in 2026 rewards candidates who can show, concretely, that they have used the relevant tools on real work, who can describe experiments they have run (including failures), and who can articulate how a workflow would change under them — not candidates who can recite credentials.

Why Frontline Hiring Is the Lane Everyone's Chasing

Frontline hiring is the biggest unsolved puzzle in enterprise recruiting, and AI vendors are circling it. Roughly 80% of the global workforce sits in frontline roles, and many of those positions stay open long enough that hiring managers are "under intense pressure to hire quickly and efficiently," per iCIMS's announcement of its Apli acquisition. The structural gap is what pulled iCIMS into a September 11, 2025 acquisition of Apli, "a fast-growing, AI-powered recruitment automation innovator focused on the frontline hiring experience." Apli runs multichannel candidate engagement through text, WhatsApp, and web chat. Its customers report the technology "automates up to 90% of the frontline hiring process, enabling 10x more hires per recruiter, reducing time to fill by up to 75% and driving 40% lower turnover on average through smarter candidate matching." Apli also "enables customers to generate up to five times more qualified candidates on average for hourly roles compared to traditional career site application flows, per customer analysis." iCIMS framed the deal as entry into a "massive, underserved opportunity," with Vera Makarov and José María Pertusa, Apli's co-CEOs, adding that "frontline hiring remains a massive, underserved opportunity." The combined product will rebrand as iCIMS Frontline AI.

The pressure to automate frontline hiring is structural on the employer side. Deloitte's 2024 Global Human Capital Trends research, summarized on Deloitte's frontline workforce management page, frames frontline workers as "the backbone of the transportation, hospitality, and services (THS) sector," with a workforce that is hourly, often entry-level, and "prone to high turnover." Deloitte's finding: "when they know their organization delivers on expectations, employees are 2.9 times more likely to go above and beyond," and "employees who believe that their organization genuinely cares about their experience and well-being are four times as likely to feel motivated to work." With frontline positions frequently unfilled and turnover high, the wedge every AI screening vendor is selling against is the time and consistency gap between posting a role and filling it.

The macro signal from Gartner (January 8, 2026) is louder still. Gartner describes hiring as an "AI arms race" with candidates and recruiters locked in tech-driven one-upmanship, and predicts that "a quarter of candidate profiles could be fake by 2028." Gartner also reports that "fewer than 1% [of layoffs linked to AI] were due to actual productivity gains," leaving CHROs "managing the fallout: skills gaps, morale issues and the risk of rehiring at higher cost." Gartner's nine future-of-work trends for 2026 include "Employers reverse the candidate fraud arms race" and "Tech-to-trades career paths blossom," with the latter citing Flexjobs data that 62% of workers "would make the switch for better pay and stability" into AI-proof skilled trades. Gartner's verdict on the differentiator: "teams redesigning workflows with AI are twice as likely to exceed revenue goals."

The labor market underneath is shifting. CNBC, reporting on May 19, 2026, found that "early-career workers in roles most exposed to AI, such as software developers, marketing professionals and sales managers, saw 16% slower growth in employment than the least exposed young workers between mid-2024 and September 2025," citing Stanford's Digital Economy Lab paper "Canaries in the Coal Mine?" CNBC also reported that AT&T "plans to invest around $38 billion over the next five years hiring and training blue-collar front-line workers, the majority of whom are skilled technicians, to expand its fiber network." AT&T "plans to hire around 3,000 technicians this year," on top of "the 10,000 the company has already hired over the last three years," with entry-level wages "between $18.18 and $31.45 per hour, depending on location and experience." Nvidia CEO Jensen Huang, at the World Economic Forum in January: "This is the largest infrastructure buildout in human history that is going to create a lot of jobs … We are going to have plumbers and electricians and construction and steel workers and network technicians and people who install and fit out the equipment." The U.S. Department of Education projects "about 2.1 million skilled trades jobs could go unfilled" by 2030, and the Associated Builders and Contractors reported a U.S. construction worker shortage of "around 350,000" this year, "expected to grow to more than 450,000 next year."

That rebalancing matters for any AI-agent vendor selling into frontline hiring: hotels, quick-service restaurants, logistics operators, and field-services contractors are the buyers most likely to pay for an automated screen that can text candidates, run a structured assessment, and hand a recruiter a ranked shortlist. Deloitte's hospitality research (January 14, 2026) reports that "81% of hoteliers prioritizing increasing employee productivity and 49% listing integrating AI-powered solutions as priority tech initiatives."

What HR Tech Observers Are Saying (Without Naming FirstWork)

Independent, named commentary on FirstWork specifically does not appear in the research file. What does appear is the broader posture of HR-tech observers and analysts toward AI-driven frontline hiring and AI-driven hiring in general.

Gartner's January 2026 briefing is the most directly applicable analyst voice in the file. Gartner's framing: "AI was supposed to make work easier. Instead, in many organizations, it's creating workslop: low-quality output that employees spend hours fixing." Gartner also flags the bidirectional nature of the fraud problem: candidates and recruiters are "locked in a cycle of tech-driven one-upmanship," with "rising fraud and falling trust." Gartner's prediction that "a quarter of candidate profiles could be fake by 2028" is the headline number from that section.

The Silicon Valley Girl interview, drawing on conversations with the named executives quoted above, surfaced three concerns that recur across the AI-hiring category:

• Fairness at the parse stage. Automated screeners that claim to evaluate "skills" still depend on the inputs they can read. A resume that uses non-standard formatting, abbreviation-heavy certifications, or scope described in project terms rather than metrics can be deprioritized before a human ever sees it. None of the named executives in the research made this exact critique on the record; it appears in the video as practitioner observation.
• Compliance exposure for conversational hiring tools. Jurisdictions like New York City's Local Law 144 and Illinois's AI Video Interview Act create audit obligations that fall on the employer using the tool, not the vendor selling it. Again, the specific regulatory framing appears in the video rather than in a published analyst note in the research file.
• Stack-replacement risk for enterprise incumbents. HR vendors with deeper enterprise footprints (HireVue, Eightfold, Pymetrics) have spent years building integrations with Workday, SAP SuccessFactors, and adjacent systems of record. A platform that recruits and hires through its own AI agents is, in effect, asking employers to route around that stack, a credible pitch to a 200-person manufacturer that never wired up Workday, and a harder sell to a Fortune 500 HR shared services organization that has spent a decade consolidating on a single system of record. The Silicon Valley Girl video frames this contrast without naming FirstWork.

Read across those three concerns, the ambient posture of HR-tech observers toward the AI-screening category in 2026 is: curious about agentic workflows, cautious on bias and compliance, and watching to see whether early vendors produce the kind of case studies (time-to-hire, quality-of-hire, pass-rate parity across demographic groups) that move the conversation from speculation to evidence.

How the AI-Hiring Category Stacks Up

The research file documents four named vendors in the AI-hiring space, with different buyers and different candidate experiences. None of the four are documented as FirstWork competitors in the research; the comparison below is grounded in what each vendor's own public materials describe.

HireVue built its name on asynchronous video interviews where candidates record answers to prompts and the platform scores delivery. The video-analysis component attracted an FTC complaint from the Electronic Privacy Information Center in 2019, and Illinois passed a law that effectively banned facial analysis in employment screening starting in 2020. HireVue walked back the visual-analysis component in 2021. What remains is closer to a natural-language processing score on the transcript itself. Candidates who clear HireVue's funnel record themselves on camera; the screen is video-first.

Pymetrics took a different angle on the same recruiter-buyer problem, building neuroscience-flavored game-based assessments (short timed tasks) and then matching candidates to roles via a model trained on the employer's existing high performers. Harver acquired Pymetrics in 2024 and folded it into a larger pre-hire assessment suite. The candidate experience is game-based rather than conversational.

Eightfold sells a Talent Intelligence Platform that builds a structured skills graph from résumés, internal HR records, and external labor-market data, then ranks candidates and recommends internal mobility paths. Eightfold's buyer is the enterprise CHRO; the candidate often does not know Eightfold is involved.

iCIMS Frontline AI (the combined iCIMS + Apli product announced September 11, 2025) is the most directly comparable product on the frontline side: conversational, mobile-first, multichannel (text, WhatsApp, web chat), with structured assessments rather than video or games. The press release positions the combined product as a single platform spanning "complex enterprise and high-volume hiring," with iCIMS Agents and iCIMS Frontline AI working together.

For a candidate, the practical takeaway from the documented category is this: each of these tools screens differently. Video-first tools score delivery; game-based tools score traits; skills-graph tools score fit against an internal model; conversational frontline tools score structured exchanges against a rubric. The research file does not document which of these approaches FirstWork uses, or whether FirstWork is a documented member of this vendor set.

What the Next Twelve Months Will Test

The structural test the research does document is industry-wide, not FirstWork-specific. Gartner (January 2026) frames the question for every CHRO: fewer than 1% of AI-linked layoffs have come from actual productivity gains, so organizations are managing the fallout of rushed AI rollouts (skills gaps, morale drag, "workslop") while still being told to adopt more AI. Frontline is the test case, per Gartner's "Tech-to-trades career paths blossom" trend: if AI screening works on a 10,000-employee hotel chain or a national retailer, the same playbook will travel to white-collar roles.

For the named executives in the research, the practical read-through is consistent. Aaron Levie's recommendation, to download Codex or Claude, wire up MCP servers, and learn how the tools are actually working under the hood, is a documented bar, not a guessed one. Yamini Rangan's three skills (scientist mindset, closeness to the ground, customer-oriented curiosity) are a documented bar. Ryan Roslansky's framing, that skills for a given role are changing 25%+ in a couple of years and projected to change 70% by 2030, is a documented signal that the credential a candidate brought to a screen two years ago is not the credential that will clear the same screen today.

For the frontline segment specifically, CNBC's reporting shows the macro rebalancing is already underway: AT&T is putting $38 billion over five years into hiring and training frontline technicians, Nvidia's Jensen Huang is publicly calling the AI buildout "the largest infrastructure buildout in human history," and the U.S. Department of Education projects 2.1 million skilled-trades jobs unfilled by 2030. Gartner's prediction of 25% fake candidate profiles by 2028 is the symmetric risk on the supply side: as the volume of frontline hiring rises, the volume of AI-generated résumés rises with it, and the screen that cannot tell the two apart is the screen that will be replaced.

First-party board data for context: ASML added 53 roles in the past seven days (salary band $31k–$235k, median $154k) and Stripe added 74 ($52k–$286k, median $238k), underscoring that the high-volume frontier-tech hiring cycle of late 2025 is broad-based. Neither company is documented in the research as a FirstWork competitor, partner, or hiring source.

Working in frontier tech? Zero G Talent tracks the openings: see every open ASML role, browse frontier tech jobs, openings at Stripe, and the people building the field.