Introducing Gem's Applicant Fraud Detection Agent: Stop fraud before it wastes your team's time

Fraudulent applications are overwhelming recruiting teams. Fake resumes, identity theft, coordinated bot attacks — these aren't just annoying. They create security threats and waste hours of recruiter time that should be spent building relationships with real candidates.

Today, we're announcing Gem's Fraud Detection Agent: an AI-powered solution that catches fraudulent applications before they reach your recruiters. Powered by Tofu, a specialized fraud detection company, the agent automatically evaluates applications using multiple fraud signals and assigns a clear risk level — with zero added friction for recruiters or candidates.

Join the waitlist to be among the first to access Fraud Detection Agent when it launches in early 2026.

What we're building

Gem's Fraud Detection Agent catches fraudulent applications before they waste your team's time or become security threats.

1. Automatic fraud detection: The agent scans every application as soon as it's submitted, evaluating risk without any manual recruiter intervention.

2. Evaluates fraud risk, not just signals: The agent evaluates multiple fraud signals together and assigns a clear risk level (high, medium, low) to each application with 90%+ accuracy. You get a complete risk assessment with transparent explanations, not just individual flags to interpret.

3. Zero added friction: Fraud detection runs automatically where you already review applications. No new tools for recruiters to learn. No ID uploads or biometric verification for candidates. 

4. Full recruiter control: Recruiters make the final decision on every candidate with full transparency. The agent flags suspicious applications and explains exactly why, but you make the final decision on whether to advance or reject.

5. Flexible deployment: Enable fraud detection per job, allowing you to control which roles require it based on risk level and application volume.

The result: recruiters focus on building relationships with real candidates while your company stays protected from security breaches and compliance risks.

How it works

Fraud Detection Agent automatically evaluates every application using multiple fraud signals:

• Resume metadata - Document properties, editing patterns, and file characteristics

• Email verification - Domain reputation, disposable email detection, and email validity

• Phone verification - Number authenticity, carrier information, and geographic consistency

• LinkedIn verification - Profile age, connection patterns, and activity history

• Person verification - Identity validation across multiple data sources

• IP and device data - Location consistency, device fingerprinting, and access patterns

[GRAPHIC: Visual showing the multiple signals being evaluated - perhaps a diagram with icons for each signal type connecting to a central "Risk Evaluation" node]

The agent evaluates all signals together using machine learning trained on billions of data points and assigns a clear risk level (high, medium, low) to each application.

Flagged applications appear in your normal review workflow, accompanied by transparent explanations of what triggered the flag. The entire process happens in seconds.

[GRAPHIC: Screenshot of flagged application with risk level and signal explanations]

Powered by Tofu: Why we partnered with a specialized fraud detection company

Building Fraud Detection Agent required a critical decision: build everything in-house or partner with a specialized company. We chose to partner with Tofu.

Tofu is a fraud detection company specializing in application-stage fraud. They've analyzed over 5 million applicants using billions of data points, with 90%+ detection accuracy across top companies, including TRM Labs, Circle.com, and Samsara. That level of specialization matters when fraudsters are constantly evolving their tactics.

We surveyed dozens of security and fraud detection vendors to find the right partner. Consistently, we heard from the market that Tofu was leading the charge in application-stage fraud detection. When we evaluated their track record ourselves, we agreed — their specialization, accuracy, and shared commitment to privacy, transparency, and recruiter control made them the clear choice.

By partnering with experts who dedicate 100% of their focus to staying ahead of new fraud tactics, we can provide our customers with the best protection available today — and ensure they remain protected as threats evolve. Tofu is SOC Type II certified and compliant with GDPR, EU AI Act, and local state laws, sharing our commitment to responsible AI and maintaining recruiter control over hiring decisions.

What makes Gem's Fraud Detection Agent different

We're not the only fraud detection solution in the market, but we took a different approach — one that prioritizes both effectiveness and experience.

Smarter detection & risk evaluation: Most solutions just flag suspicious signals and leave you to figure out what they mean. Gem provides complete risk assessments with clear levels (high, medium, low) and 90%+ accuracy.

Zero friction for recruiters and candidates: Other solutions require candidates to upload IDs and take selfies for biometric verification. Gem's fraud detection runs automatically in the background — no additional steps for candidates, no new tools for recruiters. 

Flexible pricing: We charge per application with per-job controls — you decide which roles need fraud detection based on risk and volume. We're building toward application-level control within jobs for even greater cost management. We'll share detailed pricing when we open early access next month.

Specialized partnership: We partnered with Tofu, a company specializing in fraud detection. This means you get specialized expertise, proven accuracy (90%+), and continuous improvements as fraud tactics evolve.

[GRAPHIC: Comparison table showing Gem vs. competitors on key dimensions (risk evaluation, candidate friction, pricing approach)]

The result: effective fraud protection that fits with how recruiting teams work.

Timeline and what's next

Here's how we're rolling out Fraud Detection Agent:

Now – Join the waitlist:  Sign up today to be among the first to get access. We'll keep you updated as we hit each milestone and prioritize waitlist members for early access.

February 2026 – Early access begins: We'll start enrolling select customers from the waitlist into early access. They will get hands-on time with the product, the opportunity to provide feedback that shapes the final release, and priority support from our team.

Later in Q1 2026 – General availability: Fraud Detection Agent will be available to all Gem customers. We'll share detailed pricing, documentation, and implementation guides.

Stay ahead of the fraud crisis

The companies that address this problem now will protect themselves from security threats, compliance risks, and countless wasted hours. Gem's Fraud Detection Agent gives you the tools to catch fraud early, keep recruiters focused on real candidates, and maintain the candidate experience that attracts top talent.

Join the waitlist to be among the first to access Fraud Detection Agent when it launches in early 2026.

Have questions about fraud detection or want to learn more? Reach out to your Gem account team or contact us - we're here to help.