Kolkata, Jun 17 (PTI) Axis Bank on Wednesday asserted that its digital infrastructure remains secure amid concerns over claims surrounding advanced AI models and their potential impact on banking cybersecurity, and said cyber fraud incidents largely result from customers being deceived through external communications rather than breaches in the bank's systems.

Senior bank officials said the private sector lender has maintained an unblemished data security track record in its operations and has not experienced fraud arising from weaknesses in its cybersecurity architecture.

The comments come amid heightened attention on the 'Mythos' AI model that has prompted discussions across the banking and financial services sector over its potential to accelerate the identification of security vulnerabilities.

"What we have seen, fraud incidents generally do not stem from gaps in our cybersecurity," Gaurav Gupta, Group Head, Financial Crime Intelligence, Axis Bank, said.

Gupta said it was important to distinguish between cybersecurity and cyber fraud, noting that while banks are responsible for safeguarding their digital infrastructure, many frauds occur when customers fall victim to deceptive messages, phishing attempts or other social engineering tactics.

"Cyber fraud is happening because customers are falling prey to some communication that is reaching out to them and not because of a gap in our own wall of defence," he said.

According to banking officials, the government held discussions with banks soon after claims surrounding the AI model emerged, while the RBI is also constantly carrying out supervisory exercises at the institutional level to assess preparedness against evolving cyber threats.

Sameer Shetty, Group Head – Digital Business, Transformation and Strategic Programmes, Axis Bank, described the issue as a complex one involving technology, strategy and geopolitical considerations.

"This is a complex subject. There are strategic and geopolitical dimensions to it," Shetty said.

He said the bank's board has reviewed the institution's preparedness and that a dedicated information security function continuously evaluates emerging threats and vulnerabilities.

Officials said the principal challenge posed by AI models such as Mythos is their ability to significantly speed up the discovery of weaknesses in software and digital systems.

"The thing with Mythos is that it accelerates the process of finding vulnerabilities. We did not have access to it to verify its claims," Shetty said.

To address the challenge, the bank is focusing on reducing the time required to fix identified weaknesses and deploy security updates.

"The area we are all working on is figuring out how we can accelerate patch deployment for various defects that are identified," he said.

Axis Bank has strengthened its safe banking framework over time and continues to introduce tools and AI-driven solutions to safeguard customers from fraud. Using a combination of in-house and outsourced technology teams, the bank aims to prevent fraudulent activity before it occurs.

Shetty said customer-controlled security features introduced on the bank's app, such as Mobile App Code, Face Authentication, Safety Centre and Lock FD, have helped reduce fraud incidents.

There had been a 40 per cent reduction in fraud incidents across retail mobile banking, internet banking and shopping malls in FY26 over FY25.

Meanwhile, the bank plans to open 50 new branches in West Bengal as part of its nationwide expansion programme that will add around 500 branches across India. The bank said its credit-deposit ratio stands at 70 per cent and is improving.

Axis Bank is also engaging with its in-house team, technology partners and other stakeholders to ensure that remediation efforts keep pace with advances in vulnerability detection, Shetty added. PTI BSM NN

Cybersecurity in banking hinges on distinguishing system breaches from customer deception, while faster patch deployment becomes critical against AI-driven vulnerability discovery. Cybersecurity in banking requires a distinction between internal system breaches and cyber fraud caused by customer deception through phishing and social engineering. A private sector bank said its systems remain secure and that fraud incidents have not stemmed from weaknesses in its cybersecurity architecture. It also identified the main AI-related risk as faster discovery of software vulnerabilities, making quicker patch deployment and remediation essential.