Resume fraud detection used to mean catching a fake degree. In 2026 it mostly means catching a resume that's technically plausible, written well by an assistant, and wrong about what the person can do. In a ResumeBuilder survey of 2,000 US job applicants, 24% admitted lying on their resume, most often about years of experience and skills. Engineering resumes are no exception, and remote hiring makes them harder to check.
Here's what resume fraud looks like for engineering roles now, which checks are worth the time, and why the best detection happens after the resume.
Key Takeaways
- Most resume fraud is inflation, not invention: longer tenures, bigger titles, skills used once listed as expertise.
- The organized version is worse: stolen identities and fabricated employers. US government guidance has warned about both in remote IT hiring.
- Verify employers through contact details you find yourself, not ones the candidate gives you. That's the government's own advice.
- The most reliable resume fraud detection is a conversation: ask about the specific decisions behind each claim.
What resume fraud looks like in engineering
Inflated tenure and titles. The ResumeBuilder survey found years of experience (38% of those who lied) and skills (34%) were the most common lies. In engineering, that's "senior" after two years and "led the migration" for attending its meetings.
Keyword stuffing for the screen. The resume lists 22 technologies in the skills block, matched to your job description, and the work history mentions four of them.
AI-polished work history. Clean, metric-heavy bullets that read well and say little: "Improved API performance by 40%." Whose API, measured how, compared to what? A survey reported by HR Dive found 31% of people who lied on a resume used AI to write it. That's one survey with self-reported data, but the direction isn't surprising.
Fabricated employers and identities. The organized version. The 2022 State, Treasury and FBI guidance on North Korean IT workers describes workers who present as US-based, and services have existed for years that sell references from nonexistent companies, complete with websites and phone lines.
| Fraud type | Damage if missed | Cheapest check |
|---|---|---|
| Inflated tenure or title | Medium | Ask for start and end months; compare to LinkedIn; ask what changed when they became senior |
| Skill padding | Medium | Ask for the last time they used each listed skill in production |
| AI-polished metrics | Low to medium | Ask how each number was measured and by whom |
| Fake employer | High | Find the company yourself; call its main line; check it existed at the time |
| Stolen identity | Highest | ID verification and video continuity before access |
Resume fraud detection that works
Verify employers independently. The 2022 government guidance says to verify employment and education "using contact information identified through a search engine," not information the candidate supplies. Look the company up. Call the main line. Ask HR to confirm dates and title.
Check the company existed. For small or foreign employers, check that the company had a website, a registry record or public activity during the dates claimed. A company that appeared online a month before the application is a flag.
Don't count on US databases for LATAM history. The Work Number is built on US payroll. A Wells Fargo HR page, for example, says its verifications cover only people who've worked in the US. For engineers who worked in Latin America, you'll be calling or emailing employers directly.
Interview the resume, line by line. Pick the three strongest claims. For each one: "What was the system before you got there? What did you change? What would you do differently now? Who disagreed with you?" Inflated claims run out of detail by the second question.
Look at code when it exists. Public repos and real PRs beat any bullet point, as long as you confirm they're the candidate's own work by asking them to walk you through a commit.
Where a vetted bench changes the math
Resume fraud detection is expensive because you do it per candidate, from scratch. At Ruzora, every engineer on our bench has passed a graded coding challenge and an AI interview answered on camera. Those are separate evaluations, and both are required. That doesn't make a resume true. It means the skills on it have been tested before you read it. You still interview the engineer yourself after the service agreement, with their full identity, and you should ask the resume questions above.
A Concrete Version
A 40-person insurtech is hiring a senior data engineer directly. Two finalists.
- Candidate A: 8 years, "led a Snowflake migration saving $400k a year." Asked how the $400k was measured: compute credits before and after, over two quarters, confirmed by finance. Asked what went wrong: a dbt model that doubled costs for a month. Lots of detail. The employer's main line confirms dates and title.
- Candidate B: 9 years, "architected a real-time pipeline processing 2B events a day." Asked what broke first at that scale: a general answer about partitions. Asked who disagreed with the design: no one. The employer, found by search, has a website registered seven months ago and no phone line that answers.
Time spent on verification: about 90 minutes across both. Candidate B is out. The company didn't need a fraud tool; it needed two phone calls and three good follow-up questions.
The Honest Counterpoint
Some "fraud" is just bad resume writing. Non-native English speakers often copy phrasing from templates, and plenty of honest engineers overstate a skill because recruiters filter for keywords. Treat an inflated skills list as a question, not a verdict.
Verification also has limits in Latin America. Many companies there don't have HR departments that answer reference calls from US numbers, and some startups from 2019 don't exist anymore. A failed verification call isn't proof of fraud. Combine it with the interview.
Frequently Asked Questions
How common is resume fraud?
In a 2025 ResumeBuilder survey of 2,000 US job applicants, 24% said they had lied on a resume, most often about years of experience and skills. Other surveys give different numbers depending on method, so treat any single figure as a rough guide.
What is the best resume fraud detection method?
Independent employer verification using contact details you find yourself, plus an interview that asks for the decisions behind each major claim. Tools help at volume, but the conversation catches inflation that databases miss.
Can you verify employment for engineers who worked in Latin America?
Usually by contacting the employer directly. US payroll databases generally don't cover it. International background check vendors can help, at extra cost and time. See also reference check fraud and candidate identity verification.
The Bottom Line
Resume fraud detection is mostly verification you do yourself and questions that require real memory. Call employers through numbers you find, interview the claims, and verify identity before access. If you'd rather start from engineers whose skills were tested before you met them, see our bench or read how we vet.
Roberto Espinoza is CEO of Ruzora, which helps US startups hire pre-vetted senior LATAM engineers, with a vetted shortlist in 72 hours. See available engineers.
