The AI-driven harassment interception system of the GB WhatsApp App can filter out 98.6% of spam calls. According to the 2023 tests conducted by cybersecurity company NortonLifeLock, its recognition algorithm powered by a machine learning model (trained on 470 million call records in volume) can identify harassing calls in 0.3 seconds. The false sealing rate is as low as 0.2%. Let’s consider the case of the Indian market. During 2022, on average, 9.7 spam calls were received by users daily. However, once the “Smart Blacklist” function of the GB WhatsApp App was activated, the number of complaints about harassing calls decreased by 72%, where the successful interception rate of scam calls targeting the +91 country code totaled 94%. This function also accommodates custom rules, such as blocking all international incoming calls (with + numbers) or suspicious calls with a continuous call interval of ≤15 seconds (the peak probability density is decreased by 89%).
Technically, GB WhatsApp App does two-factor authentication of number location and voiceprint recognition, which can be compared in real time with the 560 million global database of known harassing numbers (refreshed every 15 minutes), and can identify number change fraud based on voiceprint similarity analysis (with a threshold of 87%). In the 2021 telecom fraud in Brazil, the scammers made an average of 120,000 calls per day by impersonating bank customer service numbers (e.g., +5521XXXXXXXX), while the user fraud rate of GB WhatsApp App users was only 0.3% (6.8% for those who didn’t enable it). Its “Nighttime Silent Mode” can be activated to automatically reject non-contact calls between 22:00 and 7:00. On Samsung Galaxy A series phones, CPU usage only grows by 3.2%, and the peak memory usage is maintained within 18MB.
For security risks, in 2022, Check Point discovered that its interception system had a CVE-2022-44177 vulnerability. Attackers can try to produce DTMF signals of specific frequencies (e.g., 950Hz for 2 seconds) to bypass the filtering policies (with a chance of 1.1%), but the development team fixed this within 72 hours. Increase the vulnerability fix rate to 99.9%. The users are informed that enabling the “Whitelist Exception Policy” (allowing emergency numbers such as 112) can result in a false interception rate of 0.7%. The signal strength threshold needs to be manually adjusted to above -85 DBM for ensuring critical communications.
The cost-benefit analysis shows that the free anti-harassment feature of GB WhatsApp App saves the user an average of $57 per year (compared to the $4.99 per month subscription fee of Truecaller Premium). Indonesian e-commerce marketplace Tokopedia has reduced the malicious occupation rate of its customer service hotline from 34% to 6% thanks to this feature, with an average saving of 12 hours of invalid call handling costs per day. Its “pattern learning” function is able to optimize the interception model based on data manually marked by users (sample number ≥50 times), which can increase the recognition accuracy of new harassment types in the future 30 days to 93%.
Market feedback demonstrates that the feature has driven a 29% increase in the global download volume of the GB WhatsApp App in Q3 2023, with a market penetration rate of 63% in Nigeria. Sensor Tower data shows that users proactively enable the anti-harassment feature 4.2 times a day on average, and the interception volume during peak hours (20:00-22:00) accounts for 37% of the entire day. Its “group blocking” feature (blocking an entire number range at a time, e.g., +23480XXXXXXX) has reduced the success rate of transnational fraud gangs by 41%. According to the Nigerian Communications Commission report, the related criminal losses were cut down by 270 million US dollars in 2023. It is recommended for business users with a large number of calls to include the “Automatic reply Rejection” template (e.g., “In a meeting”) to increase the processing capacity of non-urgent calls by 3.8 times and reduce the average response cycle from 42 seconds to 11 seconds.