
For someone who has spent weeks applying for jobs without hearing back, an unexpected message from a recruiter can seem like a promising opportunity. A text from someone claiming to represent Indeed may lead a candidate to believe that one of their applications has finally attracted an employer’s attention. The two might spend hours discussing the position before the supposed recruiter asks the applicant to pay for a work laptop. Yet the job may never have existed, and the entire exchange may have taken place without the company’s knowledge.
Indeed requires employers to verify their identities before recruiting on its platform. In the first half of 2026, the company proactively removed more than 218,000 fraudulent employer accounts, an increase of approximately 20% from a year earlier. But scammers impersonating Indeed recruiters through text messages can bypass those safeguards entirely, leaving the company with no fraudulent account to identify or remove. That vulnerability helps explain why it began running ads on its own job board warning applicants about fraudulent offers originating elsewhere. The company emphasizes that it never contacts job seekers through text messages or social media to offer employment.
Matt Schemmel, Indeed’s vice president of trust and safety, says scammers exploit the pressure job seekers feel to respond quickly to a potential employer, particularly when they fear that delaying a reply could cost them an opportunity.
“When someone is focused on finding the next opportunity, it’s natural to want to move quickly when a promising lead comes along. Unfortunately, bad actors know this. Scammers don’t succeed because job seekers aren’t smart or cautious,” he tells me, in an exclusive interview. “They succeed because they are expert manipulators who exploit those high-stakes, vulnerable moments, often by leveraging the brand credibility of trusted names like Indeed.”
Reports Of Suspicious Accounts Roughly Tripled
One figure in the company’s recent Transparency Report stands out to Schemmel. His team removed more than 72,000 fraudulent employer accounts before a single job seeker applied, nearly twice as many as a year earlier. An application can give a scammer access to information that is difficult to recover once disclosed. A resume may reveal where someone currently works, their employment history, and a personal phone number they can use to contact them outside the platform.
“We highlight the number of accounts removed before they get an application because that exposure is what really matters. Job applications are the point where job seekers are often sharing some pretty detailed personal information, so that’s the point that we really want to know that we’re stopping bad actors,” Schemmel explained. “That metric does represent real prevention, not just catching them sometime after the fact.”
Job seekers also helped the platform identify suspicious accounts. Users reported more than 12,000, up from about 3,900 a year earlier. Schemmel attributes some of that increase to a change the company made itself.
“One of the big challenges that we’ve had in the last year has been off-platform scams. To help combat that, earlier this year we launched a proactive campaign to raise awareness including banner ads throughout the platform which made it even easier for job seekers to report jobs they are suspicious of,” he says. “That drove an increase in user reports, as expected, and I think that’s a good thing for the marketplace. We’re not always involved in the entire hiring conversation, so that partnership with job seekers to let us know that something seems off is very important.”
What Counts As An Early Catch In Fraud Detection
A hiring manager who discovers a convincing imitation of their company’s careers page may recognize the fraud immediately. But an individual employer has little visibility into whether the same scammer is impersonating other companies overseas. Indeed, which says it serves more than 3.5 million employers, has a broader view of that activity. The company monitors incoming traffic for signs of automated bots and examines patterns across employer accounts, including similarities in how job openings are posted and how applicants are contacted. Those signals help investigators identify connections between suspicious accounts and incidents that might otherwise appear unrelated to the job seekers and employers affected.
“This is the arms race happening across Security and Trust & Safety right now. Attackers have some new advantages, but I think defenders’ new advantages are even better,” Schemmel says. “One of attackers’ biggest advantages is that they can operate at scale. One of Indeed’s biggest advantages is that we can see those attacks at scale. When the same pattern is starting to show up across employers or even across countries, we’ll see it sooner than anybody.”
LinkedIn’s experience helps explain why fraudulent recruiters often try to obtain a job seeker’s phone number early in the conversation. The platform says 90% of scam messages reported by users attempted to move the interaction elsewhere, with roughly half making that request in the initial message. Those figures cover scams across LinkedIn, rather than recruitment fraud alone. In a hiring scam, however, moving the conversation to WhatsApp allows someone impersonating an employer to continue contacting an applicant even after the job board has removed the account that initiated the exchange. LinkedIn says it identified 99.5% of the fake accounts it detected before a member reported them, indicating how often its systems recognized suspicious activity without relying on user complaints.
Likewise, The Better Business Bureau offers another view of the problem through its tracking of scams across multiple categories. Online scams accounted for 53.4% of reports and 81.4% of reports involving a financial loss. Those figures illustrate how frequently fraud unfolds through digital channels, where the initial contact and eventual demand for money may occur on different services. A job board might detect the fraudulent recruitment attempt, while a messaging platform receives a complaint about a subsequent payment demand. Connecting those signals could help both services recognize a scam that neither can fully investigate from its own records alone.
AI Makes Fake Job Postings More Convincing
Generative AI has made it easier for a scammer to assemble a credible company website and produce the paperwork to match. For the company’s investigators, having a convincing job description in front of them settles very little. They can learn more from looking at how the employer behind it behaves across the site. Schemmel says AI has changed how much of that work an experienced analyst can do while a lead is still fresh.
“The number of accounts removed varies quite a bit from half to half. Some of that is better detection. We’re proud of the number we can detect and remove before job seekers apply to them. Some of it is due to the better tools available to bad actors,” he says. “Fortunately, our tools are getting better as well. Our Trust and Safety team was one of the early adopters of GenAI tools at Indeed. It’s been remarkable to see the new capabilities it gives our expert analysts. They can run down leads in hours that would’ve been a week or two of work a year ago.”
The company’s automated systems removed more than 101,000 employer accounts in the first half of 2026, close to the 100,000-plus it reported a year earlier. The intervening six months illustrate the swings Schemmel describes. Automated removals topped 529,000 in the second half of 2025.
The bigger year-over-year change came on the applicant side. Indeed rejected more than 398,000 job seeker accounts in the first half of 2026 and automatically rejected more than 343,000 of them, about 86%, up from roughly 53% a year earlier. Those totals cover a range of policy violations. An applicant who sends an abusive message and someone who invents an identity can both land in the count.
The platform employs more than 120 trust and safety professionals and works with more than 1,000 external moderation partners. Its moderators cover 27 languages across 12 countries.
“All Trust & Safety decisions require human judgment,” Schemmel says of his team’s role. “Sometimes they write narrowly targeted rules, sometimes they dig deep into borderline cases and render a specific judgment. We use AI to help our experts scale their expertise. There is always a risk that operational efficiency can reduce effectiveness. Our north star is optimizing the decision-accuracy of the whole system.”
Indeed has a commercial stake in getting that judgment right. Its Talent Scout AI tool helps employers find people to hire, while Career Scout offers job seekers AI assistance of their own. A recruiter who notices signs of AI use in an application may be looking at someone who took the job board up on its offer of help. The report gives no figures for how often the company wrongly rejected a legitimate account or how many people got one back on appeal.
When The Fake Candidate Shows Up For The Interview
Employers also face applicants who pose as someone else to get inside a company. The Justice Department says more than 130 U.S. companies unwittingly hired North Korean IT workers. For a hiring manager, the practical difficulty is that an impostor can show up for the interview. Deepfake video complicates a check as familiar as asking someone to turn on a camera.
Schemmel recommends that employers meet candidates in person when possible or keep cameras on during remote interviews so interviewers can notice technical irregularities. He expects job boards to play a larger role in verification before an applicant reaches that interview.
“We already require employers to complete rigorous verification and have started offering job seekers ways to verify their professional credentials,” he says. “Over the next couple of years, I’d expect a noticeable increase in all kinds of trust-building verification for all kinds of online activity. That’ll be particularly important for companies like Indeed that are fundamentally connecting people.”
Who handles that risk inside a company is a separate question. In a survey of 500 U.S. human resources leaders by HYPR, which sells identity security software, 53% placed responsibility for identity risk before a job offer with HR. Just 10% named the security team.
“I don’t think that enterprises are underestimating the risk in their hiring pipeline these days. A number of our larger customers (and hopefully future customers) recognize that Indeed’s central position in the marketplace gives us a unique vantage point to spot and prevent fraud in that recruiting pipeline,” he says. “Fraudulent behavior can be anything from exaggerating credentials to full-on faked identity. It’s a real concern for employers, and one that we’re focused on fixing.”
The Control Layer Verdict
Indeed can verify the employer behind a job listing, but its name appearing in a recruiter’s text message offers no such assurance. A scammer who knows an applicant’s work history, conducts a convincing interview and discusses a potential start date can make an upfront payment for a work laptop seem like a routine final step in the hiring process. Schemmel emphasizes that legitimate employers do not ask candidates to pay for work equipment in advance. The real test is whether the company’s warnings stay with job seekers long enough to influence a decision made hours later, after a seemingly credible interview, when the person requesting money has already convinced them they have a new job.


