31 December 2022

Table of Content

Leveraging voice technology to combat cyber-fraud

GS-3: Basics of Cyber Security; Money-Laundering and its prevention.

According to Reserve Bank of India (RBI) data, bank and cyber frauds have cost approximately Rs.100 crore a day over the last 7 years to India. Fraud incidents reported in 2021-2022 increased by 23.69% over the last year, albeit in a smaller amount. Increased fraud also means more losses for financial institutions and more cases for law enforcement to resolve.

Cyber security

• Cyber security is the application of technologies, processes, and controls to protect systems, networks, programs, devices and data from cyber-attacks. 
• Aim: To reduce the risk of cyber-attacks and protecting against the unauthorized exploitation of systems, networks, and technologies. It is also known as information technology (IT) security or electronic information security. 

Types of cyber threats

• The threats countered by cyber-security are three-fold –

• Cyber-crime: targeting systems for financial gain or to cause disruption.
• Cyber-attack: often involves politically motivated information gathering.
• Cyber-terrorism: disrupting electronic systems to cause panic or fear.

Methods used to threaten cyber-security

• Malware or malicious software: A software created by a cyber-criminal or hacker to damage a user’s computer. It is often spread via an unsolicited email attachment or legitimate-looking download, malware may be used by cyber criminals to make money or in politically motivated cyber-attacks. There are a number of different types of malware, including:

• Virus: A self-replicating program that attaches itself to a file and spreads throughout a computer system.
• Trojans: A type of malware that is disguised as legitimate software which when downloaded, can cause damage to or collect data from the system.
• Spyware: A program that secretly records what a user does and makes use of this information. For example, capturing credit card details.
• Ransomware: It locks down a user’s files and data, with the threat of erasing it unless a ransom is paid.
• Adware: Advertising software which can be used to spread malware

• Phishing/ social engineering: Victims are targeted with emails that appear to be from a legitimate source asking for sensitive information. Phishing attacks are often used to fool people into handing over sensitive information.
• Man-in-the-middle attack: Attack in which a cyber-criminal intercepts communication between two individuals in order to steal data. For example, intercepting data being passed from the victim’s device and the network on an insecure WiFi network.
• Denial-of-service attack: An attack in which cyber criminals prevent a computer system from fulfilling legitimate requests by trafficking the networks and servers. This renders the system unusable, preventing an organization from carrying out vital functions.
• Advanced persistent threats (APTs): In an APT, an intruder or group of intruders infiltrate a system and remain undetected for an extended period and steal sensitive data while avoiding the activation of defensive countermeasures. The recent Solar Winds breach of United States government systems is an example of an APT.

Need of the advanced solution

• The global use of smartphones has dramatically increased the use of biometrics for security, which is cumbersome, not completely secure, and prone to deepfakes. 
• Banks have traditionally relied on the use of passwords, which are the weakest link in security as 81% of hacking-related breaches involve weak passwords.
• Thus, banks and institutions seek to use voice biometric authentication to reduce the risk of fraud and identity theft.

Voice Technology (VT)

• Voice technology uses voice biometrics or voice/speech recognition technology.
• Identification: For the purpose of Identification, it uses the unique characteristics of a person’s voice. 
• Authentication: It creates a digital voiceprint and compare it with the caller’s voice for the purpose of authentication. 
• Verification: It verifies a caller swiftly in seconds by analyzing the caller’s voice and flags suspicious calls.

Features and advantages of Voice Technology

• Sensitive enough to detect if someone is impersonating the user or playing a recording
• Ability to resist playback attacks
• Identify even if the user has a cold or a sore throat
• Cheapest technology compared to other biometrics
• Does not require a reader or special device
• Non-invasive, portable and allows remote identification
• Reduce call handle times and call centre costs 
• More challenging to hack and impossible to spoof or copy
• Ensure high-accuracy authentication in seconds
• Voice authentication can provide a significant improvement in security over knowledge-based authentication methods used by fraudsters to trick people.
• Allows privacy as it does not require users to reveal personal information
• Improves user experience

Applications of Voice Technology

• In February 2019, HSBC became the first bank to introduce voice recognition for mobile banking customers, which reduces telephone banking fraud cases by over 50%.
• Helps financial institutions to ensure security for customers and employees. 
• An essential tool for forensics and law enforcement. 

• The police can leverage voice to improve investigation efficiency, identify criminals, track criminals, and better respond to and prevent crimes.

• Help airport security to check criminal background

• Face recognition technology has a high error rate and works efficiently only when the person is looking directly at the camera. 
• On contrary, Voice has a much lower error rate, and requires no eye contact. 

• Voice could be an excellent tool for the Government to disburse money for various schemes and verify the proof of life of pensioners from their homes..

Challenges 

• The Government needs to develop a mechanism for proper coordination between financial institutions and the police to investigate and prosecute fraudsters, as a fraud deterrent, and to maintain an extensive database of such criminals. 
• It is not 100% foolproof and may give false positives as it has an accuracy between 90% and 99%. 

[Ref- TH]

Fact File