The Job That Does Not Exist: Recruitment Fraud in the Age of Artificial Intelligence

By Matthew Parish

Sunday 18 October 2026

There is a particular sort of message that has become increasingly familiar on LinkedIn. It arrives from somebody whose photograph looks professional, whose biography identifies him or her as a recruitment consultant at a substantial international recruitment business, and whose profile contains just enough information to appear plausible. There is usually one curious feature: the account is relatively new, or has remarkably few connections for somebody supposedly engaged professionally in the business of knowing people.

The message itself is innocuous. Would you be interested in hearing about an opportunity? It might be a senior appointment, a consultancy, a board position or something sufficiently unusual to appeal to precisely the sort of person being approached. The purported recruiter asks whether the details might be sent by email, and there is nothing inherently suspicious about this because recruitment consultants contact people unexpectedly every day.

Then the email arrives, and the opportunity is extraordinary. It may be in precisely the country in which you have been contemplating working. It may involve exactly the combination of disciplines represented by your unusual career. The remuneration may be generous without being so fantastic as immediately to arouse disbelief. The organisation seeking to recruit may remain confidential because this is, naturally, a “retained search”. Every eccentricity in the supposed recruitment exercise has a superficially professional explanation.

Most strikingly, the job seems to understand you. That should be the moment at which you

become suspicious.

The recruiter who isn’t there

The first thing to examine is not the job description but the email address. A person claiming to conduct an executive search on behalf of a major international recruitment consultancy ought ordinarily to be corresponding from that organisation’s domain. Yet in this increasingly common variety of fraud, the message comes instead from Gmail, Outlook or another consumer email service.

The address may be constructed carefully: the recruiter’s real name, followed perhaps by the name or initials of the genuine recruitment company. To somebody reading quickly on a telephone, it looks approximately right. But approximately right is not right, particularly when the person at the other end is proposing to discuss confidential details about your professional life.

Then examine the footer. Genuine corporate recruiters generally have corporate signatures. There will ordinarily be a direct telephone number, an office address or location, a company website, regulatory or privacy language where appropriate, and other mundane details accumulated through decades of corporate bureaucracy.

Fraudulent signatures often look strangely sparse. There is a prestigious job title and an impressive company name, but no telephone number. Sometimes there is no physical office, while on other occasions the formatting merely imitates professionalism rather than exhibiting the boring complexity of the real thing. The reason may be simple: the person whose name appears at the bottom really exists, but the person writing the email is somebody else.

Recruitment consultants are unusually attractive subjects for identity theft. Their profession requires them to approach strangers, and they routinely ask for CVs. They discuss salaries, ambitions, family relocation, weaknesses in existing employment arrangements and willingness to move countries. They often operate confidentially and may communicate with candidates whom they have never met. All of this is perfectly normal when the recruiter is genuine, and therefore exceptionally useful when the recruiter is not.

The perfect job

The most sophisticated feature of the fraud is that the imaginary opportunity may be much better tailored than an ordinary recruitment approach. That is because contemporary artificial intelligence has dramatically reduced the cost of personalised deception. A person’s LinkedIn profile, published articles, professional biography, social-media output, conference appearances and other public material may provide more than enough information for a large language model to construct an imaginary position calculated to appeal specifically to him or her.

The scammer does not need to understand international arbitration, artificial intelligence, investment banking, nuclear engineering or whatever specialised profession the victim inhabits. The machine can understand enough of the vocabulary to manufacture plausibility, and it can do so with an appearance of effortless professional fluency.

This creates an interesting inversion of the traditional confidence trick. Historically, fraudulent job advertisements tended to be crude because creating a convincing fictitious professional opportunity required substantial specialist knowledge. Artificial intelligence has changed the economics, because a person who knows almost nothing about your profession can now create correspondence that sounds as though it was written by somebody who knows a great deal about it.

The resulting vacancy may therefore be suspiciously perfect. Every peculiar strand of your career suddenly turns out to be essential. Your unusual geographical experience is exactly what the mysterious client requires. Your obscure combination of professional skills, which nobody has ever previously thought belonged together, is suddenly indispensable. Your ambitions, publicly expressed in interviews or articles, happen to coincide precisely with the employer’s strategic plans.

Flattery has always been one of espionage’s cheapest currencies, and artificial intelligence has made it available industrially. The target no longer needs to be flattered by somebody who has spent weeks studying him; the preliminary research and personalised blandishments can increasingly be generated by machines.

The interview that never ends

After the victim expresses interest, something still more peculiar may happen. The recruitment process becomes increasingly elaborate while never quite advancing. The candidate is asked for a CV, then perhaps for a more detailed professional biography, and then for an explanation of particular achievements. There may follow requests for an account of leadership philosophy, preferred working environment, geographical flexibility, remuneration expectations or long-term ambitions.

More elaborate versions may request bespoke documents: an “executive alignment brief”, an “impact portfolio”, a strategic statement or some other impressive-sounding piece of recruitment paperwork. None of these expressions is necessarily evidence of fraud by itself, and genuine executive searches can involve extensive candidate materials. The warning lies in the cumulative pattern: information flows continually in one direction while verifiable evidence of an actual recruitment process never appears.

The candidate keeps providing information, while nothing concrete happens at the other end. There is no telephone conversation from a corporate number, no videoconference with an independently identifiable recruiter, no introduction to the supposed employer and no interview date. There is no independently verifiable search mandate. Instead there is another thoughtful question, followed in due course by another.

Artificial intelligence makes this sort of operation exceptionally inexpensive. The victim may believe that he is conducting a sophisticated correspondence with an experienced headhunter when in fact he is interacting with a chatbot, or with a human operator who feeds his answers into a large language model and asks it to compose the next apparently intelligent question.

The objective need not initially be money, and that distinguishes this phenomenon from familiar employment fraud. The commodity may instead be the candidate himself: his personality, aspirations, vulnerabilities, relationships and professional circumstances.

Building the dossier

A CV contains a remarkable quantity of intelligence. It identifies employers, dates, geographical movements, educational institutions, professional relationships and sometimes references. But an extended executive recruitment process can reveal much more, particularly because candidates are encouraged to present themselves candidly in order to demonstrate that they are suitable for a supposedly exceptional opportunity.

Why do you want to leave your current position? What would persuade you to relocate? How much money would change your circumstances? Which countries would you consider living in? What frustrates you professionally? What do you think you deserve that you have not received? Which institutions have disappointed you? Where do your ambitions lie? What sort of recognition matters to you?

These are recruitment questions, but they are also intelligence questions. Put together patiently, the answers produce something approaching a psychological vulnerability assessment. The target has voluntarily explained his ambitions, resentments, financial expectations, professional disappointments and preferred inducements, and he has done so because he believes that candour will improve his chances of obtaining an attractive job.

There need never be a request for a password or a bank transfer. The operation may have succeeded long before anything so vulgar occurs, because the valuable product may be a detailed understanding of the target rather than immediate access to his bank account.

This is why the possibility of state-sponsored intelligence collection deserves particular attention, although attribution must be approached carefully. Discovering that an approach is fraudulent does not, without additional evidence, establish that Russian intelligence—or any intelligence service—is behind it. Criminal fraudsters, commercial intelligence operators, stalkers, competitors and data brokers may have similar interests. Attribution in cyberspace is notoriously difficult, and accusations against a state should not be made merely because a fake recruiter has a Gmail account.

Nevertheless, the technique is obviously useful for intelligence purposes, and it fits comfortably within the broader traditions of human-intelligence cultivation. Identify a person of interest, learn his aspirations and vulnerabilities, construct an apparently accidental relationship around them, and discover what might motivate him. Artificial intelligence merely makes the preliminary cultivation vastly easier, cheaper and more scalable.

Recruitment as an intelligence operation

For opponents of Russia’s invasion of Ukraine, particularly people who travel regularly in Ukraine, work around government or military affairs, publish political material, maintain contacts across borders or occupy potentially influential professional positions, unusual recruitment approaches deserve correspondingly unusual scrutiny. Russian intelligence services have longstanding reasons to be interested in identifying the motivations, weaknesses, financial circumstances and personal networks of individuals who may possess useful access or influence.

I have encountered what appeared to be this pattern twice. That experience does not establish who was ultimately responsible, and personal experience cannot by itself prove a broader intelligence operation. But repetition teaches an important lesson: people who are publicly identifiable with causes hostile to the interests of authoritarian states should regard unsolicited professional approaches not with paranoia, but with verification.

There is an important distinction between the two. Paranoia means assuming that everybody is an intelligence officer; verification means checking whether the person who says he works for a company actually works for that company. The former is corrosive and ultimately useless, while the latter is merely sensible professional hygiene.

Fortunately, verification is remarkably easy to do. Do not verify the recruiter using the telephone number he gives you. Find the recruitment firm’s official website independently, call its published switchboard, or write to an independently obtained corporate email address. Ask whether the individual works there and whether the purported search is genuine. Search the company’s genuine staff directory where one exists. Examine the age and activity of the LinkedIn account, check whether its connections and history make sense, and compare the sender’s email domain with the company’s genuine domain.

Most importantly, perform these checks before supplying additional information. A genuine executive recruiter should not be offended by sensible verification, and recruitment businesses themselves have an interest in suppressing impersonation of their employees. Indeed, the legitimate consultant whose identity has been stolen is another victim of the operation.

The new social engineering

We are entering an era in which “phishing” is becoming an inadequate word. The classic phishing email was badly spelled, generic and absurd, and its principal defence mechanism was common sense. The new generation of artificial-intelligence-assisted social engineering can be grammatically perfect, professionally informed and individually tailored. It may continue coherently for weeks, remember what you have told it, and adapt its supposed opportunity as it learns more about you.

In other words, many of the traditional signals of fraud—bad English, implausible stories, irrelevant questions and inconsistent correspondence—are disappearing. The correspondence may instead be polished, patient and psychologically astute, precisely because generating such material has become almost costless.

A new principle is therefore required: authenticate the person, not the prose. It no longer matters very much how convincing an email sounds, because machines can produce convincing emails. It does not matter whether the recruiter appears knowledgeable about your profession, because a language model can acquire enough superficial expertise in seconds. It does not even matter whether the person knows surprisingly obscure things about you, because those things may have been harvested automatically from the internet.

What matters is whether there is an independently verifiable human being and institution behind the correspondence. Verification must therefore migrate away from linguistic clues and towards institutional facts: corporate domains, independently obtained telephone numbers, established professional histories and direct confirmation from the organisation whose name is being used.

There is a certain irony in all this. LinkedIn was constructed around the proposition that professional identity could be made visible and therefore useful. Artificial intelligence is now making visible professional identities extraordinarily easy to counterfeit. The same abundance of public information that allows genuine recruiters to find appropriate candidates allows fraudulent ones to manufacture irresistible opportunities for them.

The twenty-first-century confidence trick may therefore begin not with an African prince offering millions of dollars, but with something infinitely more persuasive: somebody offering you precisely the future you already wanted. That is why, before you send the CV, the personal statement, the portfolio of achievements and the candid explanation of everything you hope to do with the rest of your professional life, you should check the email address.