Reading

Where are the boundaries of biometrics?

  Entering the information age, the combination of biometric technology and artificial intelligence (AI) technology has produced a variety of biometric (individual) identification technologies. Nowadays, face recognition is required to enter some residential areas, which naturally caused controversy. Where is the application boundary of biometric technology? Will personal biometrics and other privacy be exposed?
  Biometric identification includes fingerprint, palm print, voice print, iris, face, gait, vein, DNA identification and other technologies. The application of these technologies should first consider safety, followed by efficiency and convenience, which is particularly important in commercial applications.
  On November 24, 2021, the United Nations adopted the first global framework agreement on the ethics of artificial intelligence – the “Recommendation on the Ethics of Artificial Intelligence” (referred to as the “Recommendation”). The “Recommendation” proposes that it is necessary to enhance the positive effects of artificial intelligence on society, but at the same time prevent the potential risks of artificial intelligence. Therefore, the Recommendation explicitly prohibits the use of artificial intelligence systems for social scoring and mass surveillance, especially facial recognition technology, which is extremely invasive and widely used.
  For security reasons, the widely used face recognition technology has been skeptical and resisted. On August 1, 2021, the Supreme People’s Court began to implement the “Regulations on Several Issues Concerning the Application of Law in the Trial of Civil Cases Related to the Use of Face Recognition Technology to Process Personal Information”, which prohibits property owners from forcing “face brushing” (using face recognition technology), etc., It shows that the application scenarios of face recognition are blocked. In comparison, fingerprint and palmprint recognition technology is superior in security. They are both efficient and less likely to cause security problems, and even if data is leaked, there is no loss of facial information that is identifiable to the naked eye. Palmprints have a higher error tolerance rate. For example, if one palmprint is wrong, it can be decoded with another palmprint, but there is only one face, and the error cannot be corrected.
  Fingerprint is the earliest technology used in biometric identification, and it became an important technology for criminal identification at the end of the 19th century. In the 1960s, new electronic computer technology entered the field of fingerprint identification, which further promoted and expanded the use of fingerprint identification. However, fingerprinting is not widely used due to the lack of features of certain individuals or groups of fingerprints, which are difficult to image, and the widespread use of fingerprints in criminal records in the past, leading some to worry that having fingerprints recorded could affect an individual’s reputation. Various venues, such as access control, cash payment, etc.
  In this case, palmprint recognition technology similar to fingerprints has begun to appear, and it may become an important biometrics technology to replace face recognition. In July 2018, Amazon used the Rekognition facial recognition system in a trial of the American Civil Liberties Union in Northern California. When the coalition used photos of congressmen for facial recognition tests, Rekognition identified 28 of them as former criminals. In July 2020, Amazon was again accused by Illinois residents of violating a state law prohibiting the unauthorized use of personal biometric data. When foreign technology giants used face recognition to defeat Maicheng, palmprint recognition gained favor.
  Since then, technology companies such as Amazon and IBM have abandoned face recognition and instead developed biometric technologies such as palm print, voice print, and iris. Amazon.com Inc. launched the Amazon One palmprint scanner in 2020, and customers can pay for certain stores by waving their palmprints on the device’s scanner. In February 2021, the company expanded its palmprint scanners to Amazon grocery stores, bookstores and four-star stores in Seattle. Two months later, Amazon.com Inc. has expanded its biometric scanning technology to brick-and-mortar stores in multiple U.S. states.
  Now, domestic technology companies have also joined the team to develop and promote the use of palmprint recognition technology. Recently, a patent for a “palmprint payment device” applied by a domestic technology company was approved. The device has two camera components and can complete payments by scanning a person’s palmprint. The device has a certain inclination angle and can be placed on the desktop, and consumers can complete palmprint recognition in a relatively comfortable posture when paying. Of course, people are also worried that palm prints collected by technology companies such as Amazon and IBM will reveal personal privacy. However, Amazon said it would optimize the technology using a “subset” of anonymized palmprint data.
  However, palmprint recognition also has some shortcomings, such as the inability to obtain rich data like fingerprint and iris scanning technology, and the reliability is slightly lower, but these situations can be improved. At present, the false rejection rate of the palmprint recognition system learned by AI is only 0.01%, which can be further improved in practice and can be fully applied to fields such as payment, attendance, transportation, and venue entry.
  Just like the diversity of individual human genes, human biological information also has the attribute of public resources. The diversity of human genes can be used in scientific research activities. Human fingerprints, palmprints, voiceprints, irises, faces and other biometric features can also be used to understand life phenomena, explore the root cause of diseases, and serve the whole society. Only by defining the public data or big data attributes of human biological information can the whole society share and use this information within the track of ethics and the rule of law, such as criminal identification and public health services. Today, data has become the five major factors of production alongside land, labor, capital, and technology. On the one hand, public data should be obtained in accordance with the law and used reasonably to transform it into productivity; on the other hand, individual privacy should also be protected.
  Currently, collecting data is easy, but protecting data security and privacy is quite difficult. To this end, some provinces and cities in China have announced the “Public Data Management Measures”, which clearly states that public data refers to administrative organs at all levels and institutions with public management and service functions that are generated or generated in the process of performing their duties and providing public services in accordance with the law. Any electronic or other records of information obtained include individual biometric information, including fingerprints, palm prints, voice prints, irises, and faces. The public data management measures stipulate that when a natural person applies to a public management and service agency for various matters requiring verification of identity information, the relevant agency shall not force the individual to use multiple methods for repeated verification or verification in a specific way; those who have verified their identity through valid identity documents , and must not be forced to repeat the verification by collecting biometric information such as fingerprints, irises, and faces, unless otherwise provided by laws and regulations or with the consent of natural persons. These regulations are consistent with the content of the Personal Information Protection Law, which came into effect on November 1, 2021, that is, no organization or individual may illegally collect, use, process, or transmit consumers’ personal information, and may not illegally buy, sell, provide or publicly consume user’s personal information.
  Under the framework of the United Nations’ Recommendation on the Ethics of Artificial Intelligence, some countries are more inclined to protect personal privacy. A US company called Clearview AI has collected a large number of face photos of British citizens through social platforms such as Facebook, and these photos are obtained without people’s knowledge. Recently, the UK Information Commissioner’s Office (ICO) warned the company that it could face a fine of up to £17 million (about 144 million yuan) for “suspected serious breaches” of UK data protection laws. Not long ago, the US state of Texas accused Meta (formerly Facebook) of capturing users’ facial recognition data billions of times without consent. State Attorney General Ken Paxton said the state, which has an estimated 20 million Facebook users, could seek billions of dollars in damages against Meta. In the face of these pressures, Meta has announced at the end of December 2021 that it will shut down its facial recognition software and delete previously collected personal facial recognition data.
  When personal biological information or other information enters the category of public data, in addition to the regulatory authorities acting in accordance with the law, individuals should also actively rely on laws and regulations to protect their rights, so as to effectively protect personal privacy and allow public data to play its productive role.

Comments Off on Where are the boundaries of biometrics?
error: Content is protected !!