blogs & articles

AI candidate fraud: What RPO buyers need to know

September 8, 2026

TL;DR

AI is making candidate misrepresentation harder to detect and pushing verification earlier in the hiring process. For RPO buyers, the challenge is not to eliminate candidate AI use, but to distinguish legitimate assistance from misrepresentation while protecting the integrity of hiring decisions. That requires layered candidate verification, the right technology and human judgment working together.

For most of recruitment’s history, verifying a candidate meant checking what they told you. AI is creating a more difficult problem: determining whether what you are seeing, hearing and assessing belongs to the candidate at all.

That changes a fairly fundamental assumption in hiring.

Candidates now have more technology at their disposal than at any point in the history of digital recruitment. Most are using it legitimately: to research employers, prepare for interviews, improve applications or communicate more effectively. But the same technology can fabricate credentials, provide undisclosed assistance during assessments, manipulate video and audio and, in more serious cases, help someone assume an identity that is not their own.

The issue for employers is therefore not whether candidates are using AI. Increasingly, they will be.

The harder question is where assistance ends and misrepresentation begins, and whether today’s hiring processes can reliably tell the difference.

The rise of AI candidate fraud in hiring

Candidate misrepresentation itself s not new. What has changed is what can now be misrepresented, and how convincingly it can be done.

A resume can contain fabricated experience. An assessment can be completed with undisclosed AI assistance. A candidate can receive real-time support during an interview. Deepfake technology can manipulate a person’s voice or appearance. Synthetic or stolen identities take the risk further still.

Research suggests employers are already confronting this problem.

The 2026 RPO Buyer Trends Report from the RPO Association and Lighthouse Research & Advisory, based on responses from 998 employers, found that 64% encounter candidate misrepresentation at least occasionally. Yet only 15% believe they can correctly identify all the ways candidates are using AI during recruitment.

That gap may matter more than either number on its own.

Employers know candidate behavior is changing. What many still lack is visibility into where acceptable AI use ends and deliberate misrepresentation begins.

Greenhouse’s 2026 AI Hiring Report points to similar concerns, with 91% of recruiters and hiring managers saying they have spotted or suspected candidate deception and 74% reporting greater concern about fake credentials than a year earlier. Gartner has also projected that by 2028, one in four candidate profiles worldwide could be fabricated.

The individual behaviors behind those numbers are not equivalent. Using AI to improve the wording of a resume is very different from presenting someone else’s experience as your own. Getting help preparing for an interview is different from receiving hidden answers during it. And neither should be confused with deliberately falsifying an identity.

Effective candidate verification depends on knowing what, exactly, needs to be verified.

Why candidate verification needs to start earlier

Background screening has traditionally taken place toward the end of recruitment. By then, an organization may already have sourced, screened, assessed and interviewed a candidate.

AI-enabled candidate misrepresentation challenges that sequence.

Consider an assessment completed with undisclosed assistance. The result may suggest a level of capability the candidate cannot independently reproduce. If that discrepancy only becomes apparent after several interview stages, the process may eventually catch it, but only after considerable time has already been invested.

Identity fraud raises the stakes further.

A video interview provides less assurance of identity when video itself can be manipulated. A polished application reveals less about communication ability when sophisticated copy can be produced in seconds. Technically strong answers may warrant more probing when candidates can access real-time AI support.

This does not mean treating every candidate who uses AI as a potential fraud risk. AI is becoming part of how people work, and hiring processes need to account for that reality.

The more useful question is whether the evidence used to make a hiring decision can be trusted.

Can the candidate demonstrate the capability described on their resume? Can they explain the work they claim to have done? Is the person being assessed the same person who will eventually join the organization?

Those questions move candidate verification beyond a final background check. Increasingly, it needs to run through the hiring process itself.

What AI candidate fraud means for RPO buyers

This shift is beginning to change what employers need from their recruitment partners.

The same RPO Buyer Trends research found that 58% of employers want their RPO provider to support fraud and risk mitigation. AI expertise has also become a growing expectation among buyers.

There is a connection between the two.

Organizations do not simply need recruitment partners that know how to use AI. They need partners that understand how AI is changing candidate behavior, where that creates risk and where additional verification is justified.

That changes some of the questions worth asking when choosing an RPO partner.

How is candidate identity established? How are credentials validated? What happens when information does not match? How is assessment integrity protected? Are recruiters equipped to recognize inconsistencies that technology may not resolve on its own? And should verification look different for a customer service hire than for someone who will have access to sensitive systems, financial data or critical infrastructure?

For higher-risk roles, these are no longer recruitment questions alone.

Cases involving stolen or fabricated identities being used to secure legitimate remote technology roles have shown how a weakness in recruitment can become a wider organizational vulnerability. Once someone has access to company systems, information or payroll, the consequences of getting identity wrong extend far beyond a poor hiring decision.

Candidate verification is therefore becoming part of a broader conversation about hiring risk.

Why technology alone cannot solve candidate fraud

It is tempting to assume that a technology-enabled problem needs a technology-led solution.

Technology certainly has a role. Identity verification, credential checks and fraud detection tools can identify inconsistencies across candidate volumes that would be difficult to manage manually.

But detection and judgment are not the same thing.

A recruiter may notice that someone with extensive experience cannot explain a project featured prominently on their resume. An answer may initially sound convincing but become inconsistent under follow-up questioning. Details shared during an interview may not align with information provided earlier.

None of these signals proves fraud on its own.

That is why interpretation matters.

The stronger model is not technology instead of recruiters, or recruiters instead of technology. It is knowing where each adds value. Technology can surface anomalies and direct attention to potential risk. Recruiters can apply context, ask better questions and determine whether an inconsistency has a reasonable explanation or warrants further verification.

That principle sits behind AMS One, bringing technology and connected data together while keeping human expertise at the center of the interactions and decisions where context and judgment matter.

As AI becomes more capable on both sides of the hiring process, knowing what to automate may only be half the challenge.

Knowing what still requires human judgment may become equally important.

How RPO providers can strengthen candidate verification

There is unlikely to be one tool or one check that solves the problem.

The behaviors are too varied, and the level of risk is not the same for every role.

A more effective approach is layered: establishing what needs to be verified, when verification should happen and how much scrutiny is appropriate for the position.

That can mean combining identity and credential checks with assessment controls, structured recruiter questioning, clear expectations around acceptable AI use and defined escalation processes when information does not add up.

It also means avoiding the opposite problem.

A recruitment process designed around suspicion can create unnecessary friction for genuine candidates. Verification needs to protect the integrity of hiring without making every applicant feel as though they are being investigated.

That balance is particularly important for RPO providers. Candidate experience has long been a measure of recruitment quality. In an environment of growing AI hiring risk, trust increasingly needs to run both ways: employers need confidence in the candidates entering their organizations, while candidates need confidence that verification is proportionate, transparent and fair.

The question for RPO buyers, then, is not simply whether a provider has fraud prevention technology. It is whether candidate integrity has been considered across sourcing, screening, assessment, interview and offer.

AMS was named a Leader in Everest Group’s 2026 Global RPO PEAK Matrix Assessment for the sixteenth consecutive year, achieving its highest-ever global position. As recruitment risks evolve, the ability to bring together technology, process and talent expertise will become an increasingly important part of what organizations expect from an RPO partnership.

Why trust is becoming an RPO outcome

RPO has traditionally been measured through outcomes employers can readily see: time to hire, cost, quality, candidate experience and access to talent.

Today’s verification challenges introduce another:confidence in the integrity of the candidate pipeline.

That does not mean eliminating AI from recruitment. Nor does it mean assuming deception whenever a candidate uses it. AI is already becoming part of how people prepare, communicate and work, just as it is becoming part of how employers source, screen and engage talent.

The distinction that matters is between assistance and misrepresentation.

Can an employer trust that the experience being presented belongs to the person presenting it? That the capability demonstrated during an assessment can be reproduced on the job? That the person interviewed is the person eventually hired?

Those questions are becoming part of what candidate quality means.

For RPO buyers, that points to a broader shift in what a recruitment partnership is expected to deliver. Speed, efficiency and access to skills will continue to matter. But moving candidates quickly through a sophisticated hiring process has limited value if the evidence behind the hiring decision cannot be trusted.

As AI changes what can be created, coached and convincingly imitated, trust in the candidate pipeline can no longer simply be assumed.

It has to be built into the process.

How confident are you in the candidates moving through your hiring process? Talk to an AMS expert about strengthening candidate verification as part of your talent acquisition strategy.

FAQs

What is AI candidate fraud?

It refers to the deliberate use of generative AI, fabricated credentials, manipulated video or audio, synthetic identities or other AI-enabled methods to misrepresent a candidate’s identity, experience or capabilities during recruitment.

How common is AI candidate fraud in 2026?

Research suggests candidate misrepresentation is already a significant concern for employers. The 2026 RPO Buyer Trends Report from the RPO Association and Lighthouse Research & Advisory found that 64% of employers encounter candidate misrepresentation at least occasionally. Greenhouse’s 2026 AI Hiring Report found that 91% of recruiters and hiring managers have spotted or suspected candidate deception.

What is deepfake hiring fraud?

Deepfake hiring fraud involves the use of AI-generated or manipulated video, audio or identity information to misrepresent who is participating in a recruitment process. This can include voice cloning, face manipulation or other techniques intended to deceive recruiters or hiring managers.

How can RPO providers help prevent candidate fraud?

RPO providers can strengthen candidate verification through a combination of identity and credential verification, appropriate fraud detection technology, assessment controls, recruiter expertise and clear escalation processes. Layered verification can help identify potential risks at different stages of recruitment rather than relying on a single check at the end.

What should RPO buyers look for in candidate verification?

RPO buyers should look beyond whether a provider has a verification tool. They should understand how identity and credentials are validated throughout recruitment, how assessment integrity is managed, how recruiters identify and escalate inconsistencies and whether verification requirements change according to the risk associated with different roles.

About AMS

AMS powers talent strategies that deliver results, redefining a new era of talent driven by people, process, data and technology.

Transform your hiring

AMS offers digital innovation and responsible AI, providing agile talent acquisition solutions and talent consulting services that can scale with your business.