1 hour ago
2
ARTICLE AD BOX
TOI analogous from Washington: In the rapidly escalating planetary contention to harness artificial quality for cyberwarfare and cyberdefense, India finds itself successful an unusually exclusive club.Anthropic, the US artificial intelligence institution down Claude AI, this week expanded entree to its highly restricted cybersecurity model, Mythos, to astir 150 organizations crossed much than 15 countries, including India, nether an inaugural called Project Glasswing. India's inclusion is notable due to the fact that the database is different dominated by Washington’s closest information partners and allies, with China a conspicuously absentee.Project Glasswing is Anthropic’s effort to code what it believes is simply a looming cybersecurity crisis. The institution says Mythos has reached a level of coding and vulnerability-discovery capableness that surpasses astir each quality information researchers. In testing, the exemplary reportedly uncovered thousands of superior bundle vulnerabilities crossed large operating systems and browsers, helping partners place and spot flaws earlier malicious actors tin exploit them.
Anthropic has been unusually cautious with Mythos. Unlike user AI products released to millions of users, Mythos remains gated due to the fact that the aforesaid capableness that tin find information weaknesses tin besides perchance beryllium utilized to exploit them. British and American information researchers person warned that frontier AI systems are approaching the constituent wherever they could importantly accelerate blase cyberattacks.That explains wherefore entree is constricted mostly to governments, captious infrastructure operators, large exertion companies, fiscal institutions and cybersecurity agencies. Participants are expected not simply to usage Mythos, but to assistance trial it, place risks and amended defenses.For India, the invitation carries strategical significance. New Delhi is not a pact state of Washington, yet it is the lone large non-allied powerfulness included successful the existent expansion.
Other countries that person entree to Mythos see France, Germany, Italy, Switzerland, the Netherlands, Spain, Belgium, Sweden, Canada, Japan, South Korea, Australia, and New Zealand.The determination to see India apt reflects a operation of geopolitical and commercialized realities. First, India has go 1 of the world's largest pools of bundle endowment and 1 of the fastest-growing markets for AI adoption.
Second, India's integer infrastructure, from banking and payments to telecom and nationalist integer platforms, has go truthful extended that vulnerabilities discovered successful Indian systems tin supply invaluable insights into securing bundle globally.This has fueled a much cynical mentation among immoderate exertion observers, 1 of whom posted connected X: Anthropic is not doing India a favour by inviting it to lunch; India is the lunch.
The statement is that India's tremendous bundle ecosystem offers a uniquely affluent situation for stress-testing frontier cybersecurity AI. By allowing Indian institutions into Project Glasswing, Anthropic gains entree to a immense scenery of real-world code, infrastructure and information challenges.
Others judge though that the narration volition beryllium mutually beneficial.Whether India is the “largest ungated AI marketplace successful the world” is harder to beryllium definitively.
While India is surely among the fastest-growing AI idiosyncratic bases globally, per-capita adoption remains little than successful wealthier countries. Still, for American AI companies seeking scale, endowment and endeavor customers, India is hard to ignore.China’s exclusion is little mysterious. Project Glasswing is fundamentally a information initiative, and Mythos is designed to observe bundle vulnerabilities astatine scale, making it a exertion with evident national-security implications.
At a clip of intensifying US-China technological rivalry, U.S lawmakers similar Senator Chris Coons ensured determination was nary accidental Washington would let specified capabilities to Chinese institutions.India besides enjoys different vantage wrong Anthropic itself. The institution has a important Indian beingness successful its enactment and engineering ranks, including Chief Technology Officer Rahul Patil, formerly CTO of Stripe.
Patil is among a increasing fig of Indian-origin executives helping signifier the adjacent procreation of frontier AI infrastructure.Anthropic has besides opened a planetary bureau successful Bengaluru, lone the 2nd aft Tokyo, tapping section engineering endowment to physique what it describes arsenic “sovereign” and localised AI infrastructure. Last October, CEO Dario Amodei met PM Narendra Modi, who is each successful connected tech, to sermon deploying AI crossed education, healthcare and agriculture, portion noting that usage of Claude Code successful India had risen five-fold successful caller months.For Indian banks and captious infrastructure operators, the improvement could yet beryllium positive. If Mythos tin place vulnerabilities earlier hostile actors do, Indian organizations specified arsenic CERT-In (Indian Computer Emergency Response Team, the authorities bureau for cybersecurity coordination) and firms specified arsenic TCS and Infosys whitethorn summation a uncommon antiaircraft caput commencement successful a satellite wherever AI is rapidly becoming some the locksmith and lockpick of cyberspace.
