Job Placement & Recruiting

CHEQ classifies each entity requesting your listings, so search engines and authorized partners can keep access while unauthorized scrapers are throttled or misdirected.

For a job site, the distinction is not human versus automated. Search engines index open roles and send candidates, and partners may carry your listings under agreement. Competitors and unauthorized aggregators collect the same listings, salary data, and company pages to rebuild what employers pay you for.

Job sites typically set a different policy for each group:

• Allow search engines that index roles and send candidates to them
• Monitor partners and aggregators operating under an agreement, so their volume stays within its terms
• Throttle or misdirect unidentified automation collecting listings and salary pages at scale

Because classification comes before the response, tightening controls on scrapers does not remove your roles from job search.

CHEQ assesses each employer account during the session, starting at signup and before the account can post a role or reach applicants.

Fake employers are a direct risk to job seekers. Scammers pose as legitimate companies, post openings that do not exist, and use the applications they receive to collect personal details.

CHEQ reassesses an employer account at the moments that matter most on a job site:

• At signup, where user data validation flags suspicious email addresses and phone numbers, and network signals can link a new account to ones you already removed
• At the first job post, where the account starts reaching applicants
• At the first resume search or export, where the account starts reaching candidate data

Reassessing trust at these transitions, rather than once at registration, is what CHEQ calls continuous affirmation. CHEQ does not review the text of a job post; it classifies the entity behind the account and passes that verdict to your trust and safety workflow.

CHEQ classifies auto-apply agents as their own entity type, so a job site can set a policy for them instead of treating them as spam bots or as people.

These tools act for real job seekers, but some submit hundreds of applications a week, and employers judge a job site by the quality of what arrives. CHEQ’s own research found AI agent activity in job placement and recruiting to be concentrated and high-intensity compared with most sectors, so this is a current policy question rather than a future one.

The policies job sites set for these agents usually cover three stages:

• Searching and reading listings, which is how an agent finds roles for its candidate and is generally low risk
• Submission pace, which can be throttled so one candidate’s agent does not flood an employer’s queue
• Submission itself, where a step-up can apply on roles whose employers want a human-completed application

Intent is an informed inference from behavior, not a certainty about any single agent. CHEQ provides the classification and its evidence, and where each line sits is your team’s call.

CHEQ assesses risk at login and again when a signed-in account does something high-value, such as exporting resumes or posting under a company name.

Recruiter accounts are a priority target on a job site because they reach candidate data in bulk. A taken-over account can export resumes and contact details, or post fraudulent jobs under a brand candidates already trust.

At login, CHEQ helps identify credential stuffing, where breached password lists are replayed against your login page. After login, classification continues, so an account that signed in normally but starts behaving unlike its owner can be stepped up or constrained.

Because responses are proportional, a recruiter signing in from a new location keeps working, and the session acting unlike that recruiter is the one that gets challenged.

CHEQ classifies the session behind each click on a sponsored listing, so what you charge employers can be checked against visits that came from job seekers.

On pay-per-click and pay-per-application models, employers are billed for engagement with their listings. When automated clicks inflate that engagement, the charges stop reflecting real candidates, and employers notice when applicant volume does not match what they paid for.

CHEQ’s reporting is built for that conversation with employers:

• Log-level detail by source, campaign, device, and geo, so invalid activity can be traced to where it came from
• Reason codes on each classification, so a billing adjustment or partner escalation has evidence behind it
• Exports through Amazon S3 and a retrospective data API, so classifications reach the billing and analytics systems you already run

The same classifications apply to your own candidate acquisition campaigns, where CHEQ supports IP blocking in Google Ads and audience exclusion across the major ad platforms.

No. CHEQ classifies the entity behind each session: what it is, whether it is authentic, and what it likely intends. It does not assess a candidate’s qualifications, verify employment history, or moderate the content of a job post.

That distinction matters in recruiting, where screening and moderation carry their own processes and obligations. CHEQ’s role sits earlier and narrower. It helps your team see whether the account, application, or request in front of them came from a real person, an AI agent, or automation, before anyone spends time on it.

CHEQ’s verdicts are designed to feed the systems that do make those decisions:

• Trust and safety review, where a flagged employer account can be checked before it posts
• Application review, where a verdict passed with the submission can mark a likely automated application before a recruiter reads it
• Analytics and billing, where invalid sessions are separated from real job seeker activity

The hiring decision, and the judgment about whether a posting is legitimate, stay with your team.