New Injection Attack Detection strengthens FaceMe’s multi-layer identity verification security to defend against emerging AI-powered impersonation attacks
Taipei, TAIWAN – August 18, 2026 – CyberLink Corp. (5203.TW), a pioneer in AI and facial recognition technologies, today announced the release of FaceMe® SDK 7.9.2, introducing Injection Attack Detection (IAD), a new security capability designed to protect mobile digital identity verification against emerging AI-powered impersonation attacks.
FaceMe® SDK combines facial recognition, Presentation Attack Detection (PAD), deepfake detection, and the new Injection Attack Detection capability into a comprehensive four-layer security architecture. Together, these technologies help organizations defend against increasingly sophisticated identity fraud while delivering more secure and trusted digital identity verification.
FaceMe Expands Protection Against a New Generation of Identity Attacks
Remote digital identity verification has traditionally focused on preventing presentation attacks using printed photos, replayed videos, or 3D masks. Today, attackers are increasingly using AI-generated media, virtual cameras, tampered applications, and manipulated device environments to bypass verification systems.
Injection Attack Detection addresses these evolving threats by detecting attempts to inject fabricated media directly into the verification workflow or compromise the integrity of the mobile application and device environment.
A Four-Layer Security Architecture for Digital Identity Verification
With the addition of Injection Attack Detection, FaceMe SDK now protects the identity verification process through four complementary technologies:
- Facial Recognition uses AI to accurately match a user's face against trusted identity records.
- Presentation Attack Detection (PAD) verifies that a live person is present and blocks spoofing attempts using photos, videos, and 3D masks.
- Deepfake Detection identifies AI-generated or manipulated facial images and videos used to impersonate legitimate users.
- Injection Attack Detection (IAD) detects virtual cameras, tampered applications, and other injection-based attacks designed to bypass normal verification workflows.
Together, these four technologies provide layered protection across image capture, liveness verification, deepfake detection, and application integrity.
Protecting the Entire Identity Verification Environment
While Presentation Attack Detection confirms that a live person is present in front of the camera, Injection Attack Detection extends protection beyond the camera itself by validating the integrity of the application, operating system, and device runtime environment throughout the verification session.
FaceMe IAD includes five core protection mechanisms:
- Image Source Protection detects virtual camera spoofing and Camera API manipulation before fabricated media reaches the verification process.
- Application Integrity Verification identifies unauthorized modifications to the application or FaceMe SDK.
- Runtime Protection detects debugging, reverse engineering, and other malicious runtime activity.
- Device Integrity Verification identifies rooted Android devices, jailbroken iOS devices, emulators, and virtual environments.
- Continuous Runtime Monitoring continuously validates device and application integrity throughout the verification session to detect suspicious changes in real time.
Together, these protections help organizations maintain the integrity of the identity verification process and reduce the risk of sophisticated impersonation attacks.
Built for High-Assurance Mobile Identity Verification
FaceMe SDK with Injection Attack Detection is designed for organizations delivering remote digital identity verification through mobile devices, including:
- Financial eKYC and customer onboarding services
- Digital banking and account recovery
- Government digital identity services
- Telecommunications customer verification
- Enterprise workforce authentication
- E-commerce identity verification
By protecting against presentation attacks, deepfakes, fabricated media, and other injection-based attacks, FaceMe SDK helps reduce the risk of fraudulent account openings, account takeover, and identity impersonation while improving trust in digital identity verification.
"As generative AI continues to evolve, organizations face increasingly sophisticated identity fraud threats," said Mei Guu, President of CyberLink Business Unit 2. "Security must extend beyond facial recognition alone. By adding Injection Attack Detection to FaceMe SDK, we're helping customers protect every stage of the identity verification process against both traditional and emerging attack techniques.”
About CyberLink
Founded in 1996, CyberLink Corp. (5203.TW) is a leading provider of digital creative editing software and artificial intelligence (AI) solutions, inclusive of facial recognition. A driver of innovation, CyberLink has embraced the transformative power of AI, harnessing it to address the demands of consumer, commercial and education markets through a wide range of solutions. From the award-winning Director Suite’s PowerDirector, PhotoDirector and Promeo, to the acclaimed AI facial recognition engine, FaceMe®, CyberLink is committed to providing exceptional experiences. Empowering creators and marketers of all skill levels, Director Suite, elevates the content creation process, while FaceMe® delivers the fast and accurate facial recognition that is key to AIoT applications for security and surveillance.
To learn more about Director Suite, visit the CyberLink website: https://www.cyberlink.com/
For more information about CyberLink’s AI facial recognition engine, visit the FaceMe® website: https://www.cyberlink.com/faceme
####
Press Contacts
