Review model
Signal-based AI analysis with readable explanations
Trust
Updated May 30, 2026Reviewed by OfferGuard AI research and review deskHow OfferGuard AI reviews a document
This page explains what the checker looks for, how verdicts are framed, and what users still need to verify manually before trusting an offer.
Signal-based review
Clear verdict bands
Manual verification still required
Verdicts
Low risk, Review, and High risk
Inputs
Text, PDFs, screenshots, and image uploads
Reviewed by
OfferGuard AI research and review desk
Proof and review
Signals, evidence, and ownership are visible on-page
Inputs
The review starts from the evidence you actually received
The scanner works best when users upload the full message trail, PDF, screenshot, or copied text so the result stays anchored to the original wording.
- +Offer text and joining letters
- +Recruiter chats and email trails
- +PDF offer letters and onboarding files
- +Payment screenshots and image uploads
Signals
Higher risk usually comes from combinations, not a single phrase
A payment request can be concerning on its own, but the verdict usually moves much faster when fee pressure is combined with weak recruiter identity, urgency, or vague role details.
- +Payment pressure plus weak company verification
- +Instant selection plus no real interview trail
- +Document pressure plus rushed joining deadlines
Case-style examples
Short examples of how risky patterns show up in real use
Offer text and recruiter chatHigh risk
Refundable fee after instant selection
The candidate was selected quickly, asked to pay a refundable onboarding fee, and kept on chat-only communication with no official company domain trail.
Next step: Pause payment and verify the employer through public channels.
Internship email screenshotReview
Mixed evidence with weak company context
The internship pitch contained rushed document pressure and vague employer details, but the uploaded evidence did not include enough official company context for a cleaner call.
Next step: Request an official company email and compare the role against a public careers page.
Offer PDF and domain matchLow risk
Lower concern but not final proof
The submitted offer aligned better with a public employer trail and did not contain fee pressure, but the result still required independent verification before signing.
Next step: Confirm the recruiter, joining path, and contract terms through official employer channels.
What the review checks
The checker extracts readable content from pasted text and uploaded files, then looks for language patterns and document signals that commonly appear in fake hiring and internship scams.
- +Pre-joining fee demands or refundable deposit language
- +UPI, wallet, or personal-bank payment instructions
- +Chat-only hiring with weak company verification
- +Urgency, pressure, and unusually fast selection claims
- +Penalty-heavy clauses, vague legal terms, or suspicious bonds
How verdicts are assigned
The output is not based on one line alone. A high-risk verdict usually appears when several strong scam markers show up together, especially payment pressure combined with identity uncertainty, unrealistic speed, or unusual document requests.
A review verdict usually means the content contains meaningful warning signs, but the evidence in the uploaded text is not strong enough to call it clearly dangerous without more verification. A low-risk verdict means fewer strong scam markers were found in the submitted evidence, not that the offer is proven safe.
What can create false comfort or false alarms
Scammers adapt their wording quickly, and some genuine employers may still use incomplete or awkward language. That means any AI checker can miss context, and some legitimate hiring flows may still look unusual at first glance.
The product is intentionally written to surface that uncertainty instead of hiding it. When the evidence is mixed, the result should push the user toward manual verification rather than blind confidence.
What users still need to verify manually
The methodology is strongest when it helps the user focus on the next checks that matter. Before acting on an offer, users should still verify the recruiter, the company domain, the interview trail, and any request for money or identity documents.
- +Check the company website and official careers page
- +Verify the recruiter email domain and sender address
- +Confirm the interview process through official channels
- +Refuse to pay unless the request can be independently validated
How methodology updates are reviewed
OfferGuard AI research and review desk reviews methodology copy, signal framing, and public trust content before major updates are published. This does not make the product infallible, but it does make the review process easier for users to audit.
FAQ
Common questions
Does a low-risk verdict mean the offer is genuine?
No. It only means fewer strong scam signals were found in the submitted content. Official verification is still required.
Why does the tool sometimes say Review instead of High risk?
A Review result means the evidence is concerning but not conclusive enough in the submitted material to call it clearly dangerous without additional checks.
Who reviews methodology updates?
OfferGuard AI research and review desk reviews methodology and public safety guidance updates.
Related guides
Keep verifying with the right next page
Verification guide
See the manual steps that should follow any AI review.
Trust center
Review data handling, payment verification, and product limitations.
Verify recruiter email domain
Use domain checks when the sender identity is uncertain.
Scam alerts
Review the combinations of signals that deserve the most caution.
Next step
Use the guide, then verify the exact evidence
These pages are designed to answer the search query directly and help users think clearly before they act. When you have the actual message, PDF, screenshot, or offer letter in hand, run the scanner and compare the result against the guidance above.
Why this page exists
We use public trust pages, visible review ownership, and related-topic links so users can verify the product itself, not just the suspicious offer they uploaded.
Publisher: DevToolStack
Legal publisher: DevToolStack
Review owner: OfferGuard AI research and review desk
Support: support@devtoolstack.in
Support target: Usually within 1 business day
Operating region: India