TECH Signal 479
The Job Board Is Lying to You
↗
Many LinkedIn job postings are either non-existent or designed to collect free training data rather than to hire.
Engineers spend time on assessments that never lead to employment, and their labor becomes proprietary data for companies that never intend to hire. This undermines trust in hiring platforms and distorts the labor market by inflating apparent demand while real opportunities shrink.
Written by elseif from the cluster below · every claim links back to a sourceThe three things worth knowing
A substantial share of listings on LinkedIn are ghost jobs with no actual headcount or funding.
Some postings lure applicants into lengthy unpaid tests whose results are used as AI training data under terms that grant the company commercial rights.
The rise of AI-driven screening and resume fraud has created a feedback loop that erodes confidence for both job seekers and employers.
THE READ
What elseif makes of it.
The market shows a surge of vague listings that look like real jobs but serve as data collection funnels. Companies post broad titles such as "AI Trainer" or "RLHF Evaluator" with no salary band or team details, attracting applicants who then face lengthy skills assessments. These assessments often take several hours and produce work that the company claims ownership of through its terms of service. The process is designed to extract effort without any intention to hire.
In many cases the entire hiring pipeline is automated: applicants record video responses, complete annotation tests, and receive rejection emails generated by language models. No human recruiter, interviewer, or evaluator is involved at any stage. Some firms go further by tracking audio, video, and browser activity during the interview session. The collected data becomes proprietary training material for the company’s AI models, while the applicant receives no compensation or feedback.
Official statistics indicate a large proportion of tech listings are phantom openings, and the share of AI-related postings has grown sharply while genuine tech hiring has declined. Employers increasingly rely on AI to auto-reject candidates, prompting job seekers to embed prompt injections in resumes to game the system. Trust has collapsed on both sides, with most hiring managers favoring AI tools and few job seekers considering the process fair, creating a self-reinforcing loop of deception and frustration.
Regulatory responses have emerged in some jurisdictions, such as Ontario’s Bill 190 and New York’s S8877, while the Texas Attorney General investigates LinkedIn for profiting from fake listings. LinkedIn has introduced a verified badge, yet many jobs remain unverified and no U.S. state has enacted enforceable law as of mid-2026. Engineers can protect themselves by reading terms of service before applying, refusing unpaid assessments longer than thirty minutes, and verifying employers through independent sources like the Better Business Bureau or Glassdoor.
Written by elseif from the cluster below · checked for specifics the sources never containedTHE CLUSTER