{ "batch": "22-identifiability", "articles": [ { "id": "learn_r22_identifiability_reverse_image_search_pimeyes", "category": "Safety", "title": "[Unverified - being checked] PimEyes: reverse image search and facial recognition", "summary": "[Unverified - being checked] PimEyes uses deep learning algorithms to analyze facial geometry from an uploaded photo and match it against publicly indexed web images.", "body": "PimEyes uses deep learning algorithms to analyze facial geometry from an uploaded photo and match it against publicly indexed web images.\n\nWhat Social Catfish actually says:\n\n“PimEyes uses deep learning algorithms to analyze the geometric structure of a face from an uploaded photograph. It generates a unique facial signature and matches it against a large database of publicly available web images.”\n\nNote: Provider states it does not identify people directly or search private social media.", "source_type": "research", "source_link": "https://socialcatfish.com/scamfish/best-reverse-image-search-tools/", "last_verified": "2026-09-02", "reliability": "STRONG PRACTICE", "source_tier": "B", "source_publisher": "Social Catfish", "evidence_quote": "PimEyes uses deep learning algorithms to analyze the geometric structure of a face from an uploaded photograph. It generates a unique facial signature and matches it against a large database of publicly available web images.", "volatility": "medium", "from_batch": "22-identifiability", "subject": "PimEyes", "topic": "reverse image search and facial recognition", "applies_to_country": "*", "applies_to_scope": "Global web crawl subject to regional blocking and terms.", "tags": { "lane": "any", "stage": "any", "region": [ "any" ], "platform": "any", "topic": [ "unclassified" ], "explicitness": "any" }, "verification": { "status": "unverified", "label": "Unverified - being checked", "reasons": [ "legal_fetch_over_cap" ], "checked_utc": "20261021-123916Z" } }, { "id": "learn_r22_identifiability_soft_biometrics_tattoos", "category": "Safety", "title": "[Unverified - being checked] Tattoo image retrieval for forensic identification", "summary": "[Unverified - being checked] Automated content-based image retrieval systems match local point descriptors of tattoo images against large law-enforcement databases.", "body": "Automated content-based image retrieval systems match local point descriptors of tattoo images against large law-enforcement databases.\n\nWhat Michigan State University / IEEE actually says:\n\n“To objectively evaluate the performance of Tattoo-ID, we constructed a database of 64,000 tattoo images provided by the Michigan State Police... Experimental results on a database of 100,000 images show that our system has excellent retrieval performance with a top-20 retrieval accuracy of 90.5%.”\n\nNote: Performance heavily relies on query image quality; low-quality queries degrade keypoint matching.", "source_type": "research", "source_link": "http://biometrics.cse.msu.edu/Publications/SoftBiometrics/LeeTongJinJain_ImageRetrievalForensics_ApplicationTattooImageDatabase_IEEEMultimedia11_b.pdf", "last_verified": "2026-09-02", "reliability": "STRONG PRACTICE", "source_tier": "B", "source_publisher": "Michigan State University / IEEE", "evidence_quote": "To objectively evaluate the performance of Tattoo-ID, we constructed a database of 64,000 tattoo images provided by the Michigan State Police... Experimental results on a database of 100,000 images show that our system has excellent retrieval performance with a top-20 retrieval accuracy of 90.5%.", "volatility": "low", "from_batch": "22-identifiability", "subject": "Michigan State Police / MSU Tattoo-ID", "topic": "tattoo image retrieval for forensic identification", "applies_to_country": "US", "applies_to_scope": "Law enforcement forensic databases.", "tags": { "lane": "any", "stage": "any", "region": [ "us" ], "platform": "any", "topic": [ "unclassified" ], "explicitness": "any" }, "verification": { "status": "failed", "label": "Unverified - being checked", "reasons": [ "legal_fetch_over_cap", "deferred_state_uncorroborated:us-mi" ], "checked_utc": "20261021-123916Z" } }, { "id": "learn_r22_identifiability_tattoo_recognition_nist", "category": "Safety", "title": "[Unverified - being checked] Tattoo Recognition Technology Challenge (Tatt-C)", "summary": "[Unverified - being checked] NIST established an open database corpus to evaluate tattoo detection and partial tattoo matching against larger gallery canvases.", "body": "NIST established an open database corpus to evaluate tattoo detection and partial tattoo matching against larger gallery canvases.\n\nWhat NIST actually says:\n\n“Given a small region from an image of a tattoo, find the larger tattoo image from which the smaller region is contained from a database. This use case has application in investigation supporting identification of an individual...”\n\nNote: Provides standardized test protocols for forensic tattoo matching.", "source_type": "official", "source_link": "https://tsapps.nist.gov/publication/get_pdf.cfm?pub_id=917896", "last_verified": "2026-09-02", "reliability": "VERIFIED RULE", "source_tier": "A", "source_publisher": "NIST", "evidence_quote": "Given a small region from an image of a tattoo, find the larger tattoo image from which the smaller region is contained from a database. This use case has application in investigation supporting identification of an individual...", "volatility": "low", "from_batch": "22-identifiability", "subject": "National Institute of Standards and Technology (NIST)", "topic": "Tattoo Recognition Technology Challenge (Tatt-C)", "applies_to_country": "US", "applies_to_scope": "Standardized evaluation corpus.", "tags": { "lane": "any", "stage": "any", "region": [ "us" ], "platform": "any", "topic": [ "unclassified" ], "explicitness": "any" }, "verification": { "status": "unverified", "label": "Unverified - being checked", "reasons": [ "legal_fetch_over_cap" ], "checked_utc": "20261021-123916Z" } }, { "id": "learn_r22_identifiability_geolocation_backgrounds", "category": "Safety", "title": "[Unverified - being checked] Bellingcat: geolocation through street layouts and background landmarks", "summary": "[Unverified - being checked] Open-source investigations geolocate visual media by mapping road layouts, architectural details, and vehicle registration markers against satellite and street-view data.", "body": "Open-source investigations geolocate visual media by mapping road layouts, architectural details, and vehicle registration markers against satellite and street-view data.\n\nWhat Bellingcat actually says:\n\n“During the video, the cameraman walks around Brega, along the roads, and this means that from watching the video we can actually map out the roads. We can see the roads, the way the curve in some areas, and where the junctions are.”\n\nNote: Relies on observable spatial relationships and environmental context.", "source_type": "research", "source_link": "https://www.bellingcat.com/resources/2014/07/09/a-beginner-guide-to-geolocation/", "last_verified": "2026-09-02", "reliability": "STRONG PRACTICE", "source_tier": "B", "source_publisher": "Bellingcat", "evidence_quote": "During the video, the cameraman walks around Brega, along the roads, and this means that from watching the video we can actually map out the roads. We can see the roads, the way the curve in some areas, and where the junctions are.", "volatility": "low", "from_batch": "22-identifiability", "subject": "Bellingcat", "topic": "geolocation through street layouts and background landmarks", "applies_to_country": "*", "applies_to_scope": "Global open-source intelligence methodology.", "tags": { "lane": "any", "stage": "any", "region": [ "any" ], "platform": "any", "topic": [ "unclassified" ], "explicitness": "any" }, "verification": { "status": "unverified", "label": "Unverified - being checked", "reasons": [ "legal_fetch_over_cap" ], "checked_utc": "20261021-123916Z" } }, { "id": "learn_r22_identifiability_fingerprint_extraction_photos", "category": "Safety", "title": "[Unverified - being checked] National Institute of Informatics (NII) Japan: fingerprint extraction from ordinary photographs", "summary": "[Unverified - being checked] Usable fingerprints can be extracted from photographs of exposed fingers taken from up to three meters away under strong lighting.", "body": "Usable fingerprints can be extracted from photographs of exposed fingers taken from up to three meters away under strong lighting.\n\nWhat BleepingComputer actually says:\n\n“Researchers from Japan's National Institute of Informatics (NII) announced yesterday they have successfully extracted usable fingerprints from photos of exposed fingers taken up to three meters away”\n\nNote: Demonstrates that visible hand details in casual selfies present biometric risks.", "source_type": "research", "source_link": "https://www.bleepingcomputer.com/news/security/scientists-extract-fingerprints-from-photos-taken-from-up-to-three-meters-away/", "last_verified": "2026-09-02", "reliability": "STRONG PRACTICE", "source_tier": "B", "source_publisher": "BleepingComputer", "evidence_quote": "Researchers from Japan's National Institute of Informatics (NII) announced yesterday they have successfully extracted usable fingerprints from photos of exposed fingers taken up to three meters away", "volatility": "low", "from_batch": "22-identifiability", "subject": "National Institute of Informatics (NII) Japan", "topic": "fingerprint extraction from ordinary photographs", "applies_to_country": "*", "applies_to_scope": "Applies wherever high-resolution images of hands are published online.", "tags": { "lane": "any", "stage": "any", "region": [ "any" ], "platform": "any", "topic": [ "unclassified" ], "explicitness": "any" }, "verification": { "status": "unverified", "label": "Unverified - being checked", "reasons": [ "legal_fetch_over_cap" ], "checked_utc": "20261021-123916Z" } }, { "id": "learn_r22_identifiability_pimeyes_optout_requires_govid_photo", "category": "Safety", "title": "[Unverified - being checked] PimEyes: Opt-out identity verification requirements", "summary": "[Unverified - being checked] PimEyes' opt-out process requires the requester to upload a photo of their face and an anonymized scan of an ID or passport for identity confirmation.", "body": "PimEyes' opt-out process requires the requester to upload a photo of their face and an anonymized scan of an ID or passport for identity confirmation.\n\nWhat PimEyes actually says:\n\n“Please upload a scan of your ID or passport for identity confirmation purposes.”\n\nNote: The opt-out page also asks for an email address for the confirmation letter; no name is requested.", "source_type": "official", "source_link": "https://pimeyes.com/en/opt-out", "last_verified": "2026-09-05", "reliability": "STRONG PRACTICE", "source_tier": "A", "source_publisher": "PimEyes", "evidence_quote": "Please upload a scan of your ID or passport for identity confirmation purposes.", "volatility": "low", "from_batch": "22-identifiability", "subject": "PimEyes", "topic": "Opt-out identity verification requirements", "applies_to_country": "*", "tags": { "lane": "any", "stage": "any", "region": [ "any" ], "platform": "any", "topic": [ "unclassified" ], "explicitness": "any" }, "verification": { "status": "unverified", "label": "Unverified - being checked", "reasons": [ "legal_fetch_over_cap" ], "checked_utc": "20261021-123916Z" } }, { "id": "learn_r22_identifiability_pimeyes_what_it_stores_face_fingerprints", "category": "Safety", "title": "[Unverified - being checked] What PimEyes stores from the open web", "summary": "[Unverified - being checked] PimEyes states the only data it stores is face fingerprints of photos found on publicly available websites, and that a user's opt-out photo will not appear in future search results.", "body": "PimEyes states the only data it stores is face fingerprints of photos found on publicly available websites, and that a user's opt-out photo will not appear in future search results.\n\nWhat PimEyes actually says:\n\n“The only data we store is face fingerprints of photos found on publicly available websites. The photo you share will not be included in future search results, and will not be used for any other purposes beyond fulfilling your opt-out request and ongoing legal obligation.”", "source_type": "official", "source_link": "https://pimeyes.com/en/opt-out", "last_verified": "2026-09-05", "reliability": "STRONG PRACTICE", "source_tier": "A", "source_publisher": "PimEyes", "evidence_quote": "The only data we store is face fingerprints of photos found on publicly available websites. The photo you share will not be included in future search results, and will not be used for any other purposes beyond fulfilling your opt-out request and ongoing legal obligation.", "volatility": "low", "from_batch": "22-identifiability", "subject": "PimEyes", "topic": "What PimEyes stores from the open web", "applies_to_country": "*", "tags": { "lane": "any", "stage": "any", "region": [ "any" ], "platform": "any", "topic": [ "unclassified" ], "explicitness": "any" }, "verification": { "status": "unverified", "label": "Unverified - being checked", "reasons": [ "thin_body", "source_unverifiable:fetch_failed:budget" ], "checked_utc": "20261021-123916Z" } }, { "id": "learn_r22_identifiability_pimeyes_no_name_attachment", "category": "Safety", "title": "[Unverified - being checked] Whether PimEyes identifies individuals by name", "summary": "[Unverified - being checked] PimEyes states its system does not identify individuals or attach names to faces, and only returns links to websites containing visually similar images.", "body": "PimEyes states its system does not identify individuals or attach names to faces, and only returns links to websites containing visually similar images.\n\nWhat PimEyes actually says:\n\n“The system does not identify individuals or attach names to faces. It returns links to websites containing visually similar images.”\n\nNote: Company statement; not independently tested for this record.", "source_type": "official", "source_link": "https://pimeyes.com/en/privacy-and-ethics", "last_verified": "2026-09-05", "reliability": "STRONG PRACTICE", "source_tier": "A", "source_publisher": "PimEyes", "evidence_quote": "The system does not identify individuals or attach names to faces. It returns links to websites containing visually similar images.", "volatility": "low", "from_batch": "22-identifiability", "subject": "PimEyes", "topic": "Whether PimEyes identifies individuals by name", "applies_to_country": "*", "tags": { "lane": "any", "stage": "any", "region": [ "any" ], "platform": "any", "topic": [ "unclassified" ], "explicitness": "any" }, "verification": { "status": "unverified", "label": "Unverified - being checked", "reasons": [ "legal_fetch_over_cap" ], "checked_utc": "20261021-123916Z" } }, { "id": "learn_r22_identifiability_pimeyes_social_media_excluded", "category": "Safety", "title": "[Unverified - being checked] PimEyes: Search coverage scope", "summary": "[Unverified - being checked] PimEyes states it does not search or index content from social media platforms, which are intentionally excluded from its search coverage.", "body": "PimEyes states it does not search or index content from social media platforms, which are intentionally excluded from its search coverage.\n\nWhat PimEyes actually says:\n\n“Social media platforms and professional networks are intentionally excluded from PimEyes' search coverage.”\n\nNote: Company statement.", "source_type": "official", "source_link": "https://pimeyes.com/en/privacy-and-ethics", "last_verified": "2026-09-05", "reliability": "STRONG PRACTICE", "source_tier": "A", "source_publisher": "PimEyes", "evidence_quote": "Social media platforms and professional networks are intentionally excluded from PimEyes' search coverage.", "volatility": "medium", "from_batch": "22-identifiability", "subject": "PimEyes", "topic": "Search coverage scope", "applies_to_country": "*", "tags": { "lane": "any", "stage": "any", "region": [ "any" ], "platform": "any", "topic": [ "unclassified" ], "explicitness": "any" }, "verification": { "status": "unverified", "label": "Unverified - being checked", "reasons": [ "legal_fetch_over_cap" ], "checked_utc": "20261021-123916Z" } }, { "id": "learn_r22_identifiability_pimeyes_no_removal_of_source_webpages", "category": "Safety", "title": "[Unverified - being checked] PimEyes: Opt-out does not delete original photos", "summary": "[Unverified - being checked] PimEyes states it does not automatically remove images from the internet, because it is a search engine that does not control content on external websites.", "body": "PimEyes states it does not automatically remove images from the internet, because it is a search engine that does not control content on external websites.\n\nWhat PimEyes actually says:\n\n“No, PimEyes does not automatically remove images from the internet.”\n\nNote: The opt-out page similarly states: 'we are not liable for the origin of the photo, nor able to take it down from the website it was posted on.'", "source_type": "official", "source_link": "https://pimeyes.com/en/privacy-and-ethics", "last_verified": "2026-09-05", "reliability": "STRONG PRACTICE", "source_tier": "A", "source_publisher": "PimEyes", "evidence_quote": "No, PimEyes does not automatically remove images from the internet.", "volatility": "low", "from_batch": "22-identifiability", "subject": "PimEyes", "topic": "Opt-out does not delete original photos", "applies_to_country": "*", "tags": { "lane": "any", "stage": "any", "region": [ "any" ], "platform": "any", "topic": [ "unclassified" ], "explicitness": "any" }, "verification": { "status": "unverified", "label": "Unverified - being checked", "reasons": [ "legal_fetch_over_cap" ], "checked_utc": "20261021-123916Z" } }, { "id": "learn_r22_identifiability_facecheck_removal_instant_hide_then_verify", "category": "Safety", "title": "[Unverified - being checked] FaceCheck.ID: Opt-out deletion process and timing", "summary": "[Unverified - being checked] FaceCheck.ID states photo removal requests are free, that photos are immediately hidden from new searches, and that within 1 to 5 business days documents are verified and the photos plus uploaded document are permanently deleted.", "body": "FaceCheck.ID states photo removal requests are free, that photos are immediately hidden from new searches, and that within 1 to 5 business days documents are verified and the photos plus uploaded document are permanently deleted.\n\nWhat FaceCheck.ID actually says:\n\n“Your photos are IMMEDIATELY HIDDEN, so they won't appear in new searches. Within 1 to 5 business days, we'll confirm your document is legitimate and the face matches. Once approved, your photos and your uploaded document are permanently deleted.”\n\nNote: Removal is approved instantly via selfie capture; an anonymized ID document instead requires manual review and a tracking link.", "source_type": "official", "source_link": "https://facecheck.id/en/RemoveMyPhotos", "last_verified": "2026-09-05", "reliability": "STRONG PRACTICE", "source_tier": "A", "source_publisher": "FaceCheck.ID", "evidence_quote": "Your photos are IMMEDIATELY HIDDEN, so they won't appear in new searches. Within 1 to 5 business days, we'll confirm your document is legitimate and the face matches. Once approved, your photos and your uploaded document are permanently deleted.", "volatility": "medium", "from_batch": "22-identifiability", "subject": "FaceCheck.ID", "topic": "Opt-out deletion process and timing", "applies_to_country": "*", "tags": { "lane": "any", "stage": "any", "region": [ "any" ], "platform": "any", "topic": [ "unclassified" ], "explicitness": "any" }, "verification": { "status": "unverified", "label": "Unverified - being checked", "reasons": [ "legal_fetch_over_cap" ], "checked_utc": "20261021-123916Z" } }, { "id": "learn_r22_identifiability_facecheck_removal_reserved_not_to_remove_criminals", "category": "Safety", "title": "[Unverified - being checked] FaceCheck.ID: Opt-out exclusions", "summary": "[Unverified - being checked] FaceCheck.ID states it reserves the right not to remove photos of child sex offenders, convicted rapists, and other violent criminals who could pose physical harm to innocent people.", "body": "FaceCheck.ID states it reserves the right not to remove photos of child sex offenders, convicted rapists, and other violent criminals who could pose physical harm to innocent people.\n\nWhat FaceCheck.ID actually says:\n\n“We reserve the right not to remove photos of child sex offenders, convicted rapists, and other violent criminals who could pose physical harm to innocent people.”", "source_type": "official", "source_link": "https://facecheck.id/en/RemoveMyPhotos", "last_verified": "2026-09-05", "reliability": "STRONG PRACTICE", "source_tier": "A", "source_publisher": "FaceCheck.ID", "evidence_quote": "We reserve the right not to remove photos of child sex offenders, convicted rapists, and other violent criminals who could pose physical harm to innocent people.", "volatility": "low", "from_batch": "22-identifiability", "subject": "FaceCheck.ID", "topic": "Opt-out exclusions", "applies_to_country": "*", "tags": { "lane": "any", "stage": "any", "region": [ "any" ], "platform": "any", "topic": [ "unclassified" ], "explicitness": "any" }, "verification": { "status": "unverified", "label": "Unverified - being checked", "reasons": [ "thin_body", "source_unverifiable:fetch_failed:budget" ], "checked_utc": "20261021-123916Z" } }, { "id": "learn_r22_identifiability_facecheck_match_score_scale_validity", "category": "Safety", "title": "[Unverified - being checked] FaceCheck.ID: Match quality score as an identifiability claim", "summary": "[Unverified - being checked] FaceCheck.ID says its match quality score ranges from 0 to 100, that 100 means the exact same photo was found, and that a score of 83 or higher indicates a good reliable match when a quality photo is uploaded.", "body": "FaceCheck.ID says its match quality score ranges from 0 to 100, that 100 means the exact same photo was found, and that a score of 83 or higher indicates a good reliable match when a quality photo is uploaded.\n\nWhat FaceCheck.ID actually says:\n\n“The match quality score ranges from 0 to 100 and indicates how closely the face matches. 100 means the exact same photo was found. A score of 83 could indicate the same person from a different angle in a different environment.”\n\nNote: No peer-reviewed accuracy data is published on this page; the vendor also warns that 'a computer can't establish a guaranteed identity match based on low-quality internet photos.'", "source_type": "official", "source_link": "https://facecheck.id/en/Face-Search/FAQ", "last_verified": "2026-09-05", "reliability": "STRONG PRACTICE", "source_tier": "A", "source_publisher": "FaceCheck.ID", "evidence_quote": "The match quality score ranges from 0 to 100 and indicates how closely the face matches. 100 means the exact same photo was found. A score of 83 could indicate the same person from a different angle in a different environment.", "volatility": "medium", "from_batch": "22-identifiability", "subject": "FaceCheck.ID", "topic": "Match quality score as an identifiability claim", "applies_to_country": "*", "applies_to_scope": "testable: upload a known-quality photo and inspect the returned score; thresholds are self-reported by the vendor and sh", "tags": { "lane": "any", "stage": "any", "region": [ "any" ], "platform": "any", "topic": [ "unclassified" ], "explicitness": "any" }, "verification": { "status": "unverified", "label": "Unverified - being checked", "reasons": [ "legal_fetch_over_cap" ], "checked_utc": "20261021-123916Z" } }, { "id": "learn_r22_identifiability_facecheck_no_identity_output", "category": "Safety", "title": "[Unverified - being checked] Whether FaceCheck identifies people", "summary": "[Unverified - being checked] FaceCheck.ID states it does not identify people but instead identifies websites that have displayed images resembling the person in the uploaded photo.", "body": "FaceCheck.ID states it does not identify people but instead identifies websites that have displayed images resembling the person in the uploaded photo.\n\nWhat FaceCheck.ID actually says:\n\n“FaceCheck does not identify people. Instead, the tool identifies websites that have displayed images resembling the person in the uploaded photo at some point in time.”", "source_type": "official", "source_link": "https://facecheck.id/en/Face-Search/FAQ", "last_verified": "2026-09-05", "reliability": "STRONG PRACTICE", "source_tier": "A", "source_publisher": "FaceCheck.ID", "evidence_quote": "FaceCheck does not identify people. Instead, the tool identifies websites that have displayed images resembling the person in the uploaded photo at some point in time.", "volatility": "low", "from_batch": "22-identifiability", "subject": "FaceCheck.ID", "topic": "Whether FaceCheck identifies people", "applies_to_country": "*", "tags": { "lane": "any", "stage": "any", "region": [ "any" ], "platform": "any", "topic": [ "unclassified" ], "explicitness": "any" }, "verification": { "status": "unverified", "label": "Unverified - being checked", "reasons": [ "thin_body", "source_unverifiable:fetch_failed:budget" ], "checked_utc": "20261021-123916Z" } }, { "id": "learn_r22_identifiability_facecheck_data_thumbnails_and_urls", "category": "Safety", "title": "[Unverified - being checked] What FaceCheck stores in its index", "summary": "[Unverified - being checked] FaceCheck.ID states it collects and caches only a tiny low-resolution thumbnail of a face and a URL link to a website, and does not store the original source images or personally identifiable information.", "body": "FaceCheck.ID states it collects and caches only a tiny low-resolution thumbnail of a face and a URL link to a website, and does not store the original source images or personally identifiable information.\n\nWhat FaceCheck.ID actually says:\n\n“FaceCheck collects and caches only a tiny low-resolution thumbnail of a face and a URL link to a website. We do not store the original source images.”\n\nNote: Company statement; the FAQ also says searched images are not added to the database.", "source_type": "official", "source_link": "https://facecheck.id/en/Face-Search/FAQ", "last_verified": "2026-09-05", "reliability": "STRONG PRACTICE", "source_tier": "A", "source_publisher": "FaceCheck.ID", "evidence_quote": "FaceCheck collects and caches only a tiny low-resolution thumbnail of a face and a URL link to a website. We do not store the original source images.", "volatility": "low", "from_batch": "22-identifiability", "subject": "FaceCheck.ID", "topic": "What FaceCheck stores in its index", "applies_to_country": "*", "tags": { "lane": "any", "stage": "any", "region": [ "any" ], "platform": "any", "topic": [ "unclassified" ], "explicitness": "any" }, "verification": { "status": "unverified", "label": "Unverified - being checked", "reasons": [ "legal_fetch_over_cap" ], "checked_utc": "20261021-123916Z" } }, { "id": "learn_r22_identifiability_facecheck_search_history_retention", "category": "Safety", "title": "[Unverified - being checked] FaceCheck.ID: Retention of user search data", "summary": "[Unverified - being checked] FaceCheck.ID states it does not log search history or IP addresses, and that a search is temporarily stored and deleted after 24 hours.", "body": "FaceCheck.ID states it does not log search history or IP addresses, and that a search is temporarily stored and deleted after 24 hours.\n\nWhat FaceCheck.ID actually says:\n\n“We do not log your search history or your IP address. Uploaded photos are deleted. They are not indexed. Your search is temporarily stored for 24 hours. After 24 hours, the search is deleted.”\n\nNote: Company statement, not independently verified.", "source_type": "official", "source_link": "https://facecheck.id/en/Face-Search/FAQ", "last_verified": "2026-09-05", "reliability": "STRONG PRACTICE", "source_tier": "A", "source_publisher": "FaceCheck.ID", "evidence_quote": "We do not log your search history or your IP address. Uploaded photos are deleted. They are not indexed. Your search is temporarily stored for 24 hours. After 24 hours, the search is deleted.", "volatility": "low", "from_batch": "22-identifiability", "subject": "FaceCheck.ID", "topic": "Retention of user search data", "applies_to_country": "*", "tags": { "lane": "any", "stage": "any", "region": [ "any" ], "platform": "any", "topic": [ "unclassified" ], "explicitness": "any" }, "verification": { "status": "unverified", "label": "Unverified - being checked", "reasons": [ "legal_fetch_over_cap" ], "checked_utc": "20261021-123916Z" } }, { "id": "learn_r22_identifiability_googlelens_similar_images_ranking", "category": "Safety", "title": "[Unverified - being checked] How Lens finds visually similar images", "summary": "[Unverified - being checked] Google states Lens compares objects in a picture to other images and ranks those images by similarity and relevance to the objects in the original picture.", "body": "Google states Lens compares objects in a picture to other images and ranks those images by similarity and relevance to the objects in the original picture.\n\nWhat Google actually says:\n\n“Lens compares objects in your picture to other images, and ranks those images based on their similarity and relevance to the objects in the original picture.”", "source_type": "official", "source_link": "https://lens.google/howlensworks/", "last_verified": "2026-09-05", "reliability": "VERIFIED RULE", "source_tier": "A", "source_publisher": "Google", "evidence_quote": "Lens compares objects in your picture to other images, and ranks those images based on their similarity and relevance to the objects in the original picture.", "volatility": "low", "from_batch": "22-identifiability", "subject": "Google Lens", "topic": "How Lens finds visually similar images", "applies_to_country": "*", "applies_to_scope": "testable: reverse search a distinctive photo and check whether visually similar images are returned.", "tags": { "lane": "any", "stage": "any", "region": [ "any" ], "platform": "any", "topic": [ "unclassified" ], "explicitness": "any" }, "verification": { "status": "unverified", "label": "Unverified - being checked", "reasons": [ "thin_body", "source_unverifiable:fetch_failed:budget" ], "checked_utc": "20261021-123916Z" } }, { "id": "learn_r22_identifiability_googlelens_metadata_as_ranking_signal", "category": "Safety", "title": "[Unverified - being checked] Metadata as an identifiability signal in Lens", "summary": "[Unverified - being checked] Google states Lens may use signals such as words, language, and metadata on an image's host site to determine ranking and relevance of results.", "body": "Google states Lens may use signals such as words, language, and metadata on an image's host site to determine ranking and relevance of results.\n\nWhat Google actually says:\n\n“Lens may also use other helpful signals, such as words, language, and other metadata on the image's host site, to determine ranking and relevance.”\n\nNote: Confirms Lens consumes host-site metadata, which can tie an image to pages carrying an identity.", "source_type": "official", "source_link": "https://lens.google/howlensworks/", "last_verified": "2026-09-05", "reliability": "VERIFIED RULE", "source_tier": "A", "source_publisher": "Google", "evidence_quote": "Lens may also use other helpful signals, such as words, language, and other metadata on the image's host site, to determine ranking and relevance.", "volatility": "low", "from_batch": "22-identifiability", "subject": "Google Lens", "topic": "Metadata as an identifiability signal in Lens", "applies_to_country": "*", "tags": { "lane": "any", "stage": "any", "region": [ "any" ], "platform": "any", "topic": [ "unclassified" ], "explicitness": "any" }, "verification": { "status": "unverified", "label": "Unverified - being checked", "reasons": [ "legal_fetch_over_cap" ], "checked_utc": "20261021-123916Z" } }, { "id": "learn_r22_identifiability_yandex_ris_cbir_visual_phrases", "category": "Safety", "title": "[Unverified - being checked] How Yandex reverse image search matches images", "summary": "[Unverified - being checked] Yandex states its reverse image search (named 'Sibir') converts submitted images into sets of 'visual phrases', numerical representations of key features, then selects images containing the same visual phrases from billions of pictures on the internet.", "body": "Yandex states its reverse image search (named 'Sibir') converts submitted images into sets of 'visual phrases', numerical representations of key features, then selects images containing the same visual phrases from billions of pictures on the internet.\n\nWhat Yandex actually says:\n\n“The technology transforms pictures, submitted by users in the form of search queries, into sets of “visual phrases” – that is, numerical representations of images’ key features. Next, from the billions of pictures available on the internet, the search engine selects images containing the same visual phrases as those in the submitted picture, sifting out all the rest.”\n\nNote: From Yandex's official company blog.", "source_type": "official", "source_link": "https://yandex.com/company/blog/57", "last_verified": "2026-09-05", "reliability": "STRONG PRACTICE", "source_tier": "A", "source_publisher": "Yandex", "evidence_quote": "The technology transforms pictures, submitted by users in the form of search queries, into sets of “visual phrases” – that is, numerical representations of images’ key features. Next, from the billions of pictures available on the internet, the search engine selects images containing the same visual phrases as those in the submitted picture, sifting out all the rest.", "volatility": "low", "from_batch": "22-identifiability", "subject": "Yandex", "topic": "How Yandex reverse image search matches images", "applies_to_country": "*", "applies_to_scope": "testable: upload a source photo to yandex.com/images and compare returned matches against the described ranking system.", "tags": { "lane": "any", "stage": "any", "region": [ "any" ], "platform": "any", "topic": [ "unclassified" ], "explicitness": "any" }, "verification": { "status": "unverified", "label": "Unverified - being checked", "reasons": [ "thin_body", "source_unverifiable:fetch_failed:budget" ], "checked_utc": "20261021-123916Z" } }, { "id": "learn_r22_identifiability_yandex_ris_use_cases_identity", "category": "Safety", "title": "[Unverified - being checked] Yandex's stated use cases for reverse image search", "summary": "[Unverified - being checked] Yandex lists detecting false identities on social networks or dating websites and checking copyright infringement as two of the five most frequent uses of its reverse image search.", "body": "Yandex lists detecting false identities on social networks or dating websites and checking copyright infringement as two of the five most frequent uses of its reverse image search.\n\nWhat Yandex actually says:\n\n““Are they lying to me with a false image?” Is that a genuine portrait of the person trying to make friends with you on a social network or dating website?”\n\nNote: Relevance: reverse image search is positioned explicitly as an identifiability/verification tool.", "source_type": "official", "source_link": "https://yandex.com/company/blog/57", "last_verified": "2026-09-05", "reliability": "STRONG PRACTICE", "source_tier": "A", "source_publisher": "Yandex", "evidence_quote": "“Are they lying to me with a false image?” Is that a genuine portrait of the person trying to make friends with you on a social network or dating website?", "volatility": "low", "from_batch": "22-identifiability", "subject": "Yandex", "topic": "Yandex's stated use cases for reverse image search", "applies_to_country": "*", "tags": { "lane": "any", "stage": "any", "region": [ "any" ], "platform": "any", "topic": [ "unclassified" ], "explicitness": "any" }, "verification": { "status": "unverified", "label": "Unverified - being checked", "reasons": [ "legal_fetch_over_cap" ], "checked_utc": "20261021-123916Z" } }, { "id": "learn_r22_identifiability_exif_captures_when_where_how", "category": "Safety", "title": "[Unverified - being checked] What EXIF metadata records and how it travels", "summary": "[Unverified - being checked] Consumer Reports reports that details about when, where and how a photo was taken are captured automatically by smartphones and digital cameras, stored as EXIF data, and that the EXIF data travels with the photo.", "body": "Consumer Reports reports that details about when, where and how a photo was taken are captured automatically by smartphones and digital cameras, stored as EXIF data, and that the EXIF data travels with the photo.\n\nWhat Consumer Reports actually says:\n\n“Details about when, where, and how a photo was taken are captured automatically by smartphones and digital cameras, and stored as Exif (Exchangeable Image File Format) data. Information on everything from exposure settings to altitude may be included. And the Exif data travels with the photo—from the camera to your hard drive or a website.”\n\nNote: Consumer Reports testing/editorial on a Flickr photo showed exact date, time and GPS location recoverable from EXIF.", "source_type": "research", "source_link": "https://www.consumerreports.org/electronics-computers/privacy/what-can-you-tell-from-photo-exif-data-a2386546443/", "last_verified": "2026-09-05", "reliability": "STRONG PRACTICE", "source_tier": "B", "source_publisher": "Consumer Reports", "evidence_quote": "Details about when, where, and how a photo was taken are captured automatically by smartphones and digital cameras, and stored as Exif (Exchangeable Image File Format) data. Information on everything from exposure settings to altitude may be included. And the Exif data travels with the photo—from the camera to your hard drive or a website.", "volatility": "low", "from_batch": "22-identifiability", "subject": "EXIF metadata", "topic": "What EXIF metadata records and how it travels", "applies_to_country": "*", "applies_to_scope": "testable: inspect a smartphone photo's EXIF with Properties/Inspector and confirm fields such as GPS and capture time.", "tags": { "lane": "any", "stage": "any", "region": [ "any" ], "platform": "any", "topic": [ "unclassified" ], "explicitness": "any" }, "verification": { "status": "unverified", "label": "Unverified - being checked", "reasons": [ "legal_fetch_over_cap" ], "checked_utc": "20261021-123916Z" } }, { "id": "learn_r22_identifiability_exif_platforms_that_strip", "category": "Safety", "title": "[Unverified - being checked] Which platforms strip EXIF on upload", "summary": "[Unverified - being checked] Consumer Reports reports that Craigslist, Facebook, Imgur, Instagram, Twitter and WhatsApp remove EXIF data from images shown to others when users upload them.", "body": "Consumer Reports reports that Craigslist, Facebook, Imgur, Instagram, Twitter and WhatsApp remove EXIF data from images shown to others when users upload them.\n\nWhat Consumer Reports actually says:\n\n“Many other sites that let you display photos to friends or the public remove the Exif data before the images are shown to others. If you upload pictures to Craigslist, Facebook, Imgur, Instagram, Twitter, or WhatsApp, the Exif data won't be available to the people who see them.”\n\nNote: Platform policies change; treat as a 2019 snapshot.", "source_type": "research", "source_link": "https://www.consumerreports.org/electronics-computers/privacy/what-can-you-tell-from-photo-exif-data-a2386546443/", "last_verified": "2026-09-05", "reliability": "STRONG PRACTICE", "source_tier": "B", "source_publisher": "Consumer Reports", "evidence_quote": "Many other sites that let you display photos to friends or the public remove the Exif data before the images are shown to others. If you upload pictures to Craigslist, Facebook, Imgur, Instagram, Twitter, or WhatsApp, the Exif data won't be available to the people who see them.", "volatility": "high", "from_batch": "22-identifiability", "subject": "EXIF metadata", "topic": "Which platforms strip EXIF on upload", "applies_to_country": "*", "applies_to_scope": "list is as published in 2019 and may have changed; the article also notes email and texted photos typically keep EXIF.", "tags": { "lane": "any", "stage": "any", "region": [ "any" ], "platform": "any", "topic": [ "unclassified" ], "explicitness": "any" }, "verification": { "status": "unverified", "label": "Unverified - being checked", "reasons": [ "legal_fetch_over_cap" ], "checked_utc": "20261021-123916Z" } }, { "id": "learn_r22_identifiability_exif_social_platforms_still_collect", "category": "Safety", "title": "[Unverified - being checked] Whether platforms collect EXIF even when not displayed", "summary": "[Unverified - being checked] Consumer Reports reports that Facebook told it Facebook does collect and process EXIF data, including device make/model, camera settings and date the photo was taken.", "body": "Consumer Reports reports that Facebook told it Facebook does collect and process EXIF data, including device make/model, camera settings and date the photo was taken.\n\nWhat Consumer Reports actually says:\n\n“However, a Facebook representative said via email that Facebook does collect and process Exif data, including “information like the make and model of the device used to take the photo, the camera settings, and the date the photo was taken . . . to make your experience better and to keep people safe.””\n\nNote: Attributed statement to a Facebook representative.", "source_type": "research", "source_link": "https://www.consumerreports.org/electronics-computers/privacy/what-can-you-tell-from-photo-exif-data-a2386546443/", "last_verified": "2026-09-05", "reliability": "STRONG PRACTICE", "source_tier": "B", "source_publisher": "Consumer Reports", "evidence_quote": "However, a Facebook representative said via email that Facebook does collect and process Exif data, including “information like the make and model of the device used to take the photo, the camera settings, and the date the photo was taken . . . to make your experience better and to keep people safe.”", "volatility": "medium", "from_batch": "22-identifiability", "subject": "EXIF metadata", "topic": "Whether platforms collect EXIF even when not displayed", "applies_to_country": "*", "tags": { "lane": "any", "stage": "any", "region": [ "any" ], "platform": "any", "topic": [ "unclassified" ], "explicitness": "any" }, "verification": { "status": "unverified", "label": "Unverified - being checked", "reasons": [ "legal_fetch_over_cap" ], "checked_utc": "20261021-123916Z" } }, { "id": "learn_r22_identifiability_background_clues_identify_creator", "category": "Safety", "title": "[Unverified - being checked] Photo backgrounds exposing a creator's location", "summary": "[Unverified - being checked] Unbias the News reports that an OnlyFans creator suspected that people who doxxed her found her address by analysing her YouTube videos and figuring out her location based on the background visible in shots of her balcony.", "body": "Unbias the News reports that an OnlyFans creator suspected that people who doxxed her found her address by analysing her YouTube videos and figuring out her location based on the background visible in shots of her balcony.\n\nWhat Unbias The News actually says:\n\n“She suspects that they found her address by combing through the information published about her on Celebforum, analysing her YouTube videos, pausing on shots of her balcony — and figuring out her location based on the background.”\n\nNote: Part of a cross-border investigation by four reporters; interviews with 32 creators.", "source_type": "research", "source_link": "https://unbiasthenews.org/consent-violated-how-creators-pay-the-price-for-stolen-onlyfans-content/", "last_verified": "2026-09-05", "reliability": "STRONG PRACTICE", "source_tier": "B", "source_publisher": "Unbias The News", "evidence_quote": "She suspects that they found her address by combing through the information published about her on Celebforum, analysing her YouTube videos, pausing on shots of her balcony — and figuring out her location based on the background.", "volatility": "low", "from_batch": "22-identifiability", "subject": "Backgrounds as identifiers", "topic": "Photo backgrounds exposing a creator's location", "applies_to_country": "*", "applies_to_scope": "Anecdotal case report; not an experimental result. The creator's suspicion is reported, not confirmed by the outlet.", "tags": { "lane": "any", "stage": "any", "region": [ "any" ], "platform": "any", "topic": [ "unclassified" ], "explicitness": "any" }, "verification": { "status": "unverified", "label": "Unverified - being checked", "reasons": [ "legal_fetch_over_cap" ], "checked_utc": "20261021-123916Z" } }, { "id": "learn_r22_identifiability_yandex_similar_images_help_center", "category": "Safety", "title": "[Unverified - being checked] Yandex Images similar-image search", "summary": "[Unverified - being checked] Yandex's own support centre states that Yandex Images can help users find images similar to an input image.", "body": "Yandex's own support centre states that Yandex Images can help users find images similar to an input image.\n\nWhat Yandex actually says:\n\n“Yandex Images can help you find images similar to an input image. This might be useful if, for example, you're searching for a particular product based on visual features.”", "source_type": "official", "source_link": "https://yandex.com/support/images/similar.html", "last_verified": "2026-09-05", "reliability": "STRONG PRACTICE", "source_tier": "A", "source_publisher": "Yandex", "evidence_quote": "Yandex Images can help you find images similar to an input image. This might be useful if, for example, you're searching for a particular product based on visual features.", "volatility": "low", "from_batch": "22-identifiability", "subject": "Yandex", "topic": "Yandex Images similar-image search", "applies_to_country": "*", "applies_to_scope": "testable: click 'Similar' on any image in Yandex Images search results.", "tags": { "lane": "any", "stage": "any", "region": [ "any" ], "platform": "any", "topic": [ "unclassified" ], "explicitness": "any" }, "verification": { "status": "unverified", "label": "Unverified - being checked", "reasons": [ "legal_fetch_over_cap" ], "checked_utc": "20261021-123916Z" } }, { "id": "learn_r22_identifiability_tattoo_police_identification_case", "category": "Safety", "title": "[Unverified - being checked] Tattoos as identifiers: Documented case of a person identified via a tattoo", "summary": "[Unverified - being checked] The Independent reported that investigators identified Capitol riot suspect James Burton McGrew using a giant 'King James' stomach tattoo that was revealed when he lifted his shirt in police bodycam footage.", "body": "The Independent reported that investigators identified Capitol riot suspect James Burton McGrew using a giant 'King James' stomach tattoo that was revealed when he lifted his shirt in police bodycam footage.\n\nWhat The Independent actually says:\n\n“Investigators used police bodycam footage from the Capitol riot allegedly showing James Burton McGrew lifting his shirt to cover his face, revealing a tattoo across his belly that reads \"King James\" to identify him.”\n\nNote: Clarified by other outlets' reporting on the same FBI complaint that the tattoo matched a 2012 booking photo; this record only quotes what The Independent states. Testable in the sense that multiple outlets reported the same underlying FBI court filing.", "source_type": "research", "source_link": "https://www.the-independent.com/news/world/americas/us-politics/tattoo-capital-rioter-arrested-video-b1858464.html", "last_verified": "2026-09-05", "reliability": "STRONG PRACTICE", "source_tier": "B", "source_publisher": "The Independent", "evidence_quote": "Investigators used police bodycam footage from the Capitol riot allegedly showing James Burton McGrew lifting his shirt to cover his face, revealing a tattoo across his belly that reads \"King James\" to identify him.", "volatility": "low", "from_batch": "22-identifiability", "subject": "Tattoos as identifiers", "topic": "Documented case of a person identified via a tattoo", "applies_to_country": "US", "applies_to_scope": "Documented single case; identifies how tattoos function as an identification signal in law enforcement.", "tags": { "lane": "any", "stage": "any", "region": [ "us" ], "platform": "any", "topic": [ "unclassified" ], "explicitness": "any" }, "verification": { "status": "unverified", "label": "Unverified - being checked", "reasons": [ "legal_fetch_over_cap" ], "checked_utc": "20261021-123916Z" } } ] }