AI headshot services now cost $10-30 and produce results indistinguishable from professional photography. LinkedIn is flooded with them — some from real people improving their profile photos, some from fake personas that don't exist. Here's the distinction that matters for recruiters, and how to check any LinkedIn photo in under three seconds.
Most AI headshots on LinkedIn today are AI-enhanced — a real person uploaded 15-30 casual photos of themselves to a service like Aragon, ProfilePicture.ai, or ProShot, and received back a polished, professional-quality headshot. The face is real. The person exists. The lighting, background, and framing were improved by AI. This is legal, consistent with LinkedIn's terms (the photo is a likeness of you), and increasingly common among job seekers who can't afford $200-500 for a professional photographer.
A much smaller but higher-risk category is AI-generated headshots — a face produced entirely from scratch by a generative AI model, with no real person as input. This is what fake profile operators use. The profile exists purely to scam, impersonate, or commit fraud. The face has never existed.
For a recruiter or hiring manager, these two cases have completely different implications. An AI-enhanced headshot of a real candidate is a cosmetic concern at most. An AI-generated headshot of a fake persona is fraud.
| AI-enhanced headshot | AI-generated headshot | |
|---|---|---|
| Real person behind profile? | Yes | No |
| Face in photo exists? | Yes — real person | No — fabricated |
| Passes AI detector? | Often — face is real | Usually fails |
| Shows up on reverse image search? | Sometimes — if person has online presence | No — never existed |
| LinkedIn ToS compliant? | Yes — face is a likeness | No — misrepresentation |
| Risk to recruiter | Cosmetic concern only | Potential fraud / fake applicant |
The major AI headshot services — Aragon, ProfilePicture.ai, ProShot, Secta, HeadshotPro — all operate on the same model: the user uploads 15-30 real photos of themselves taken in various lighting and angles. The service fine-tunes a diffusion model on those photos, then generates headshots in professional settings. The face in the output is recognizably the real person — the AI improved the photo, it didn't invent a new person.
This is fundamentally different from using Midjourney to prompt "professional LinkedIn headshot of a 35-year-old male executive" and getting back a face that doesn't correspond to anyone. The AI headshot services require real input photos of the real person; face-generation tools don't.
From a detection standpoint: AI-enhanced headshots from Aragon et al. often pass AI image detectors because the underlying face data is real — the detector finds few purely synthetic signals. Fully fabricated faces from Midjourney, StyleGAN, or similar tools typically do trigger detectors because no real face data anchored the generation.
On LinkedIn's web interface in Chrome:
Note: AI-enhanced headshots from legitimate services sometimes trigger the AI detector even when the person is real. Use the detector as one signal among several, not as a definitive verdict. See: Why real photos sometimes get flagged as AI.
A suspicious LinkedIn headshot is the starting point for verification, not the end. The profile-level signals that matter most for fake applicant detection go well beyond the photo — see the full guide at Fake AI job applicants: a hiring manager's guide. Gartner projected in November 2024 that 1 in 4 job applicants will be fake by 2028 as AI-generated personas become cheaper and easier to operate at scale.
For high-stakes hiring (roles with financial access, executive positions, security clearance), the photo check is worth running on every external candidate. For volume hiring, a spot-check on candidates who advance past initial screening is a reasonable risk-balanced approach.
Not reliably by visual inspection. Modern AI headshot services produce results indistinguishable from professional photography. A recruiter cannot confidently identify an AI-generated headshot without using a detection tool. Right-click the photo in Chrome and run FauxSpy — it takes three seconds and returns a probability score. Combine with a Google Image Search for strongest signal.
LinkedIn requires profile photos to be a likeness of you. An AI headshot service that uses real photos of you to generate a polished version is compliant — the face is yours. A fabricated AI face with no real person behind it violates LinkedIn's terms and constitutes identity misrepresentation. The policy distinction is whether the face in the photo actually belongs to the account holder.
AI-enhanced: a real person's photos → AI generates a professional version preserving their real face. AI-generated: no real person → AI creates a face from scratch. Services like Aragon and ProfilePicture.ai produce AI-enhanced headshots (real person, improved photo). Tools like Midjourney used without any reference photo produce AI-generated faces (nobody behind them). The first type is common and generally harmless; the second is used in fraud.
In Chrome: right-click the photo → "Check with Faux Spy" for an AI detection score. Also right-click → "Search image with Google" to check for prior online presence. A fabricated AI face returns no image search results and typically triggers the AI detector. A real person's AI-enhanced photo often passes the detector and may appear on their other profiles. Use both signals together.