New Delhi Telecom service providers should be shielded from any liability even if legitimate communications are inadvertently intercepted by artificial intelligence or machine translation (AI-ML) systems, taking into account the Telecom Regulatory Authority of India (TRAI) Was designed to block unwanted calls and messages. Mathan Babu, chief information security officer and data privacy officer at Vodafone Idea, said in an interview that telecom companies were mandated to deploy such technologies to prevent spamming.
India’s third largest carrier has used AI-ML to design a solution to block spam without directly deploying an AI-ML engine. It said it will take some more time to fully deploy the AI-ML system as telcos are expecting some initial hurdles such as accidental filtering of authentic messages falsely identified as spam.
“Deployment of any specific AI/ML model in live requires validation against large data sets and is still a work in progress. Models have a high probability of blocking true positives and any AI/ML Any decision has to be carefully evaluated and can be taken forward only if there is immunity from both the owner of the data set and the regulator,” Babu said.
Even as telcos implement AI algorithms to thwart unwanted communications, the demand for such measures is on the rise, especially after the rise in fraudulent spam messages on WhatsApp. TRAI directs telcos to use AI to help millions of Indian users. Spammers are exploiting the Meta-owned messaging platform to defraud unsuspecting users by calling them from international numbers.
According to the Local Circles survey published on Wednesday, 46% of Indian WhatsApp users received unwanted audio and video calls, which has increased in the past month. It said that 59% of the calls originated from unknown international numbers.
In a previous attempt to block unwanted communications, the regulator had, in March 2021, mandated telecom companies to use blockchain or distributed ledger technology, but this resulted in massive disruption of financial and UPI transactions as the scrubbing mechanism failed. blocked one-time messages. -password or OTP TRAI had to temporarily suspend the mandate but later resumed it as the threat of spam messages and calls remained.
Babu added that Vodafone Idea already has methods and processes in place that have already shown better results even without implementing AI/ML models and has shown that it can better control and protect its own customer base. Is capable of receiving spam messages using deployed DLTs. technology. He said the solution was shown to TRAI and some banks.
He said the AI model, which learns and evolves from the inputs provided to it, could potentially identify marketing messages being sent by, for example, a bank, to customers as spam messages and potentially Can block them, thus getting in the way. marketing efforts.
He added that in such cases, feedback is needed to close the analysis from the marketer, who is the real messenger on an ongoing basis, to improve the model and deploy it effectively.
“AI/ML models take longer to mature to provide the desired level of accuracy,” he said.
catch all technology news And updates on Live Mint. download mint news app to receive daily market update & Live business News,