TypingDNA enables easy identity verification for web apps with typing biometrics
TypingDNA has developed technology that expands the limited biometric authentication options not requiring specialized sensors or advanced hardware. Its approach works with existing keyboards, and can be used to protect any app by matching a short typing sample. TypingDNA’s cybersecurity SaaS uses artificial intelligence to match two or more typing patterns, which consist of “press and flight times” keyboard input.
Typing patterns are virtually impossible to steal, and are also simpler to use for web applications than other biometrics, CMO Cristian Tamas told Biometric Update in an interview. “If you would like to verify a person’s identity with another biometric, you would have to have special permission in the browser to request verification data from the hardware, so it gets very complicated.”
The Romania-based company recently secured a funding round from Gecad Ventures and a group of angel investors, after participating in the local MVP Academy accelerator. The deal also brought Radu Georgescu, who founded RAV Antivirus and exited after Microsft acquired it in 2003, to the company as an investor. TypingDNA previously developed its product through participation in Deutsche Telekom’s Hub:raum and the Microsoft Accelerator Bootcamp.
A huge potential market for TypingDNA is eLearning. Online educational courses like MOOCs can be taken by hundreds of thousands of users, making it nearly impossible to verify the identity of each student without using biometrics. Traditional proctoring is not effective, and integrating other biometrics imposes hardware requirements, such as a webcam or fingerprint scanner, in addition to the challenges of capturing and verifying images from them. Because typing data is automatically collected by browsers, special permissions are not required as they are for other methods, and implementation and authorization are privacy-friendly, more natural, and less restrictive.
eLearning providers can implement TypingDNA’s API to increase the value of the certification they provide by verifying the identity of the student, and ensuring that the person being certified is the same one who took the exam. It also helps prevent lost revenue from account sharing.
TypingDNA can also improve password recovery processes, which often rely on SMS messages that quite literally play into the hands of a malicious actor with a stolen device, or “security questions” that hackers can potentially find the answers to online. Using typing biometrics as an authentication layer enhances the security of these automated processes without annoying the end user.
Tamas also sees major potential for identity management platforms to supplement or replace weaker factors in two-factor authentication, to enhance the security of corporate apps, and for the banking, finance, and payments industry, which is required to adopt two-factor authentication in Europe as PSD2 comes into effect.
By using its AI algorithm to analyze press and flight times, TypingDNA allows the application to authenticate based on any text, giving its clients the flexibility, for instance, to provide a promotional message as sample text to authenticate the user’s typing behaviour.
TypingDNA is also bringing a continuous authentication solution to market to minimize human device security errors by cutting off access to a PC or Mac laptop or desktop as soon as the typing biometric score falls below a specified threshold.
With two-factor authentication growing, typing biometrics could be the most appropriate form of “what you are” to enable it for many uses. “It’s not yet widely known, but it’s gaining traction,” Tamas said.
Global Industry Analysts Inc. predicts the typing biometrics market will approach $800 million by 2020. Continued innovation of products and its core AI algorithms position TypingDNA to meet increasingly rigorous demands as the market for typing biometrics takes off.