FM Sitharaman flags AI cyber risks amid Anthropic ‘Mythos’ concerns

Union concern curate Nirmala Sitharaman connected Thursday urged banks to fortify cybersecurity systems and follow a much proactive attack to tackle emerging threats linked to precocious artificial quality models, peculiarly amid concerns implicit Anthropic’s “Claude Mythos”.Chairing a high-level gathering with slope chiefs alongside IT curate Ashwini Vaishnaw, Sitharaman directed lenders to instrumentality “all indispensable pre-emptive measures” to unafraid IT systems, support lawsuit information and safeguard fiscal resources.

Push for real-time menace sharing

The concern ministry stressed the request for stronger coordination crossed institutions. “It was advised that a robust mechanics for real-time menace quality sharing whitethorn beryllium established among banks, @IndianCERT and different applicable agencies,” the ministry said successful a station connected X.Banks person besides been asked to instantly study suspicious enactment to CERT-In and support adjacent coordination with applicable authorities.

‘New and little understood’ AI threat

Sitharaman warned that AI-driven risks disagree from accepted cyber threats and necessitate much precocious responses.Speaking aft the meeting, she said, “What we have… proved ourselves to beryllium that we are protective, mightiness not beryllium enough. We request thing acold much versatile,” arsenic quoted by ANI.

She noted that portion Indian banks person maintained a beardown cybersecurity grounds owed to continuous upgrades and digitisation, the evolving quality of AI threats demands caller strategies and tools.

Coordinated effect led by banks

The curate asked the Indian Banks’ Association to make a coordinated organization mechanism, with the SBI president acceptable to pb efforts. Banks volition collaborate to place vulnerabilities, measure investments and research however AI itself tin beryllium utilized to antagonistic specified risks.She besides urged lenders to onboard apical cybersecurity professionals and specialised agencies to fortify monitoring and defence systems.

Concerns astir ‘Claude Mythos’

The gathering assumes value amid planetary attraction connected Anthropic’s AI exemplary “Claude Mythos”, which reportedly has the capableness to place and exploit bundle vulnerabilities astatine scale. The exemplary has demonstrated the quality to uncover thousands of bugs, including decades-old flaws successful large systems.Anthropic has restricted nationalist merchandise of the model, citing “unprecedented cybersecurity risks” and imaginable misuse.A elder authoritative said Indian systems stay unafraid and determination is nary request for undue concern, though the authorities and RBI are intimately studying imaginable risks, reported PTI.Sitharaman emphasised that vigilance and preparedness volition beryllium key, with banks expected to heighten coordination and readiness to woody with what she described arsenic an “unprecedented” technological challenge.