The emergence of advanced AI systems such as Claude Mythos from Anthropic is prompting telecom operators to reassess cybersecurity preparedness, as these tools uncover software vulnerabilities that traditional systems often miss, senior executives indicated.
Bharti Airtel’s Chief Technology Officer Randeep Sekhon said the company is in discussions with suppliers to assess what needs fixing in light of these AI-driven insights.
“We are talking to our suppliers and vendors to understand what vulnerabilities could be exposed and what needs to be fixed,” Sekhon said while speaking to the press. He indicated that operators are working with their technology partners to evaluate and plug such gaps, given that much of the telecom software stack is vendor-owned.
Vendor Collaboration
The focus is on understanding the nature of these vulnerabilities and ensuring timely fixes through suppliers, executives indicated. Meanwhile, Vodafone Idea’s CEO Abhijit Kishore said that the telco is actively working to strengthen network resilience and security frameworks, while evaluating similar concerns through its internal initiatives. Although he did not confirm any specific tie-ups.
However, the bright spot is that despite the heightened attention, executives said that there is no immediate cause for alarm. The vulnerabilities being flagged are mostly incremental software bugs rather than systemic threats to network infrastructure. On being asked if the government has approached the telcos on similar concerns, Sekhon indicated ‘not yet’. This comes at the backdrop of Finance Minister Nirmala Sitharaman chairing a high-level meeting with the banking sector on Thursday, to discuss potential risks and prepare defensive measures.
Balancing Defensive Readiness
Since telecom networks rely heavily on external vendors for core systems, operators depend on companies such as Nokia and Ericsson to address any identified issues. The role of operators, therefore, is to coordinate and ensure patches and safeguards are implemented swiftly.
Sekhon added that much of what is currently labelled as AI, such as conversational tools replacing traditional search engines remains an incremental improvement, not a fundamental shift. Most changes are expected from real-time inferencing, where networks can dynamically respond to changing conditions.
Separately, Airtel acknowledged that operators are seeing a rise in memory-related costs, affecting server-heavy deployments. However, the impact on overall network spending remains limited, as not all components rely on memory-intensive architecture.