Fake Text Message Investigation Report
Report generated: Dec 2, 2025
image_scan/PNweGOaQmhuWdA3zDzluxZjuaDaoRTLjJPS4KeG5.png
https://webvetted.com/storage/image_scan/PNweGOaQmhuWdA3zDzluxZjuaDaoRTLjJPS4KeG5.pngWhy we think so?
Flags present: the image contains text that automated tools mark as synthetic across many text boxes, and OCR pulled a US phone number (+1 (626) 327-7245) twice. Reverse-image checks show the picture appears across multiple Instagram/YouTube posts (copy detected). There are no QR codes or links and the moderation scan found no violent or hate content. Overall, this looks suspicious — treat the phone contact and any instructions in the image with caution. ⚠️
Risk Insights
Synthetic text detected
- Many OCR boxes labeled 'text-artificial' with ~0.96–0.98 confidence.
- Synthetic text undermines claims of authenticity and makes messages easier to spoof.
- Treat any instructions or contact details in the image as unverified.
No QR or links found
- No immediate drive-by redirect or malicious URL in the image.
- Risk is via contact (phone) or social-engineering rather than embedded links.
Category Scores
Red Flags & Warnings
- High-confidence detection of synthetic/AI-generated text on multiple text boxes.
- Presence of an extracted phone number (potential contact method for scams).
- Image appears repeatedly across public social posts, suggesting reuse rather than a single authoritative source.
Detailed Checks & Insights
0-100 Scale
AI / manipulated text detection (image OCR)
Score: 15
AI / manipulated text detection (image OCR)
"Sightengine found numerous bounding boxes classed 'text-artificial' (probabilities ~0.96–0.98) and reported has_artificial=0.98 while has_natural≈0.001; this pattern strongly suggests the visible text was generated or overlaid rather than captured from a natural printed source."
Reason: Automated OCR labeled many text regions as 'text-artificial' with very high probabilities (≥0.95), indicating the text was likely synthetic or composited.
Extracted text scanned for scam keywords or URLs
Score: 40
Extracted text scanned for scam keywords or URLs
"Sightengine's text extraction returned a US-format phone number twice and no link or domain; while a number alone isn't proof of fraud, it gives a direct contact vector commonly used in scam flows."
Reason: OCR produced a clear phone number (+1 (626) 327-7245) but no URLs; phone numbers can be used to contact or social-engineer targets.
OpenAI moderation for harmful or manipulative content
Score: 90
OpenAI moderation for harmful or manipulative content
"The omni-moderation model returned flagged=false and all major category scores near zero, meaning the image does not contain content typically blocked for safety reasons."
Reason: No moderation categories (hate, sexual-minors, illicit, violent) were triggered; image content is not overtly abusive or illegal.
QR code or embedded link destination reputation
Score: 85
QR code or embedded link destination reputation
"Sightengine's qr and link detections returned empty arrays for personal/link/social/spam/blacklist fields, so there is no direct clickable target in the image to inspect."
Reason: No QR codes or embedded links were detected in the image, reducing immediate risk from drive-by redirects.
Reverse image search (reuse / provenance)
Score: 50
Reverse image search (reuse / provenance)
"Lens results include multiple Instagram posts, YouTube channels, and other social links (copy_detected=true). Broad reuse makes it harder to trust the image as an original authoritative asset."
Reason: Google Lens shows many matches across Instagram and YouTube; the image appears reused and not uniquely tied to an official organization.
Your Next Steps
-
1
Do not call or text the extracted phone number unless you can verify it independently from a trusted website or official directory.
-
2
If you received this image claiming to be from a company or government agency, contact the organization through an address or phone number listed on their official site — not the one in the image.
-
3
If you must interact, forward the image to a trusted friend or your institution’s security team and ask them to verify its origin before acting.
-
4
Keep a screenshot of the message and any related communication in case you need to report it to your local consumer protection agency or platform support.
Evidence & Citations
-
Sightengine image analysis (OCR, AI-text, QR, metadata)
OCR detected multiple 'text-artificial' boxes with very high probabilities (~0.95–0.98) and extracted the phone number +1 (626) 327-7245 twice; no QR/link detected; image sharpness and brightness are high.
-
Google Lens reverse-image results (multiple social-media matches)
Lens returned many Instagram/YouTube/other social matches and reported copy_detected=true, indicating the image appears across public posts rather than being unique.
-
OpenAI Omni moderation check
Moderation model flagged the image as not violating categories (harassment, hate, sexual-minors, illicit, violence all false), with only negligible category scores.
🕵🏻 Keep vetting
Upload or link to an image to test for AI or tampering.
How WebVetted investigates fraud »Understand our methodology, founder background, and why people trust us.
Scam Encyclopedia »Browse platform-specific scams / red flags and see real-world encounters curated by our analysts.