#Japan #Tokyo #AI Data #Deepfake #NII

The Fight Against DeepFakes: A Collaborative Initiative

The rise of DeepFake technology has transformed from a niche domain accessible only to tech experts to a significant concern that anyone with access to technology can exploit. This shift has led to serious issues, including misinformation on social media, identity theft, child exploitation, damage to corporate reputations, and even financial fraud. In light of these escalating threats, AI Data Corporation, based in Minato, Tokyo, is embarking on a new collaborative initiative with the National Institute of Informatics (NII), aiming to eliminate the scourge of DeepFakes from society.

Addressing the DeepFake Challenge

The initiative, dubbed the “Zero DeepFake Society,” approaches the challenge from multiple fronts, recognizing the urgent need for a comprehensive strategy that transcends various sectors, including law enforcement, education, and private enterprises. With insights garnered from their experience assisting police, local governments, and corporations in digital forensics, AI Data is turning its focus to developing a robust model for detecting and mitigating DeepFakes. Central to this effort is NII’s advanced AI technology, known as “SYNTHETIQ VISION,” which provides a foundational platform for their proposed solutions.

Three Pillars of the Collaborative Model

AI Data’s vision for achieving a DeepFake-free society operates on three primary pillars:

1. Law Enforcement: The company envisions applying this technology by testing its efficacy within police departments, cybercrime units, and prosecutors’ offices. Potential use cases include assessing the effectiveness of initial screenings for child exploitation materials, verifying the authenticity of videos posted on social media, and evaluating the role of AI in substantiating evidence against cyber fraud.

2. Educational Institutions and Local Governments: The initiative also aims to protect students from DeepFake threats, such as synthetic bullying videos. This includes checking educational materials for DeepFake content and developing training resources for students and educators to enhance media literacy and discernment skills.

3. Private Sector: Involving corporations in financial services, media, e-commerce, and public relations, the strategy focuses on areas like bolstering identity verification processes in electronic Know Your Customer (eKYC) applications and enhancing measures against the infiltration of fakes in brand materials and advertisements.

These points represent a selection of the possible applications under consideration. As practical implementations are further evaluated and discussions with relevant entities continue, AI Data remains committed to refining this model.

The Core Technology: SYNTHETIQ VISION

SYNTHETIQ VISION is an advanced AI developed by NII that is specifically tailored for detecting DeepFake content. Its key features include:

• - Analytical capabilities to identify subtle artifacts unique to generated or synthesized facial visuals, outputting potential authenticity scores.
• - A design that allows for evaluation of challenging visuals that defy easy human identification, including images that have undergone compression or other processing.
• - An architecture that stores outcomes and related logs for future validation and auditing.

AI Data positions SYNTHETIQ VISION as a core component of its initiative, focusing on how effectively it can detect and counteract DeepFake production across various domains and how it can integrate seamlessly with existing investigative and auditing processes.

Designing the Model for Implementation

AI Data is currently exploring several potential models for offering this new technology:

• - SaaS Model: A system where users upload videos to analyze scores indicative of potential fakeness in real time.
• - API Integration: Interfaces that allow current platforms and systems to adopt this technology.
• - OEM Offering: Methods to integrate DeepFake detection capabilities into other companies’ products.
• - Secure Deployment: Evaluations of configurations tailored for government, financial institutions, and local authorities within protected environments.

A Call for Collaborative Action Towards a Zero DeepFake Society

This initiative does not merely focus on technology; it addresses a structural societal issue that requires a collective response. AI Data plans to work alongside law enforcement, educational bodies, financial institutions, and various stakeholders to ensure that the model proposed can be implemented effectively for social good.

Through joint verification, guidance on operational protocols, and integration with existing digital forensic services, AI Data aims to foster a robust environment for combating DeepFakes, thereby realizing the goal of a Zero DeepFake Society gradually.

AI Data Corporation has a proven track record in safeguarding and maximizing data assets over its long history, reinforcing its commitment to innovation through this ambitious undertaking.

About AI Data Corporation

AI Data Corporation, established in April 2015, has built its business on safeguarding data infrastructures and intellectual properties for over 20 years. With a capital of 1 billion JPY, it has earned the trust of more than 10,000 companies and 1 million individuals.