AI Dispatch | Agentic AI Moves Into Indian Finance

News

1. AI Dispatch | Agentic AI Moves Into Indian Finance

2. Anthropic Discloses Fourth Claude Breach Missed in Earlier Review

3. Visa, Mastercard, Ant Push Common Standards for AI Agents Making Payments

4. OpenAI Calls for Mandatory US Frontier AI Safety Rules

5. NPCI Launches Two Agentic AI Platforms

6. SBI Chief Urges ‘Know Your Agent’ Checks as AI Moves From Advice to Execution

AI and payment connectivity dominated India’s fintech agenda this week. Prime Minister Narendra Modi called for more overseas UPI links, NPCI launched two agentic AI platforms and Paytm expanded its enterprise AI push. Elsewhere, concerns over advanced AI intensified while Google backed 16 climate-focused AI projects across Asia-Pacific.

Modi Pushes for More Global UPI Links

Prime Minister Narendra Modi called for India to connect UPI with payment systems in more countries, prioritizing markets with large Indian communities and strong trade links. He cited Singapore’s UPI-PayNow connection as a model and urged fintech companies to expand beyond payments while using technologies including agentic AI.

Read more: At Fintech Fest, PM Pushes for Global UPI Links to Cut Costs for Indians Abroad

NPCI Launches Two Agentic AI Platforms

NPCI launched AiNxt and AtOM, two agentic AI platforms aimed at simplifying AI development and payment-network operations. AiNxt lets organizations build AI agents using their own models through no-code, low-code and developer tools, while AtOM automates integration, partner onboarding, testing and certification. NPCI has released both platforms as open source.

Read more: NPCI Launches Two AI Platforms for Payments

Paytm Pushes Into Enterprise AI Agents

Paytm is expanding into enterprise AI through Pi, or Paytm Intelligence, a platform built around agents that can handle tasks such as customer acquisition, payments and service. Bloomberg reported that Paytm plans to offer the technology to banks, lenders, insurers and other financial institutions in India and the UAE.

Read more: Paytm Plans AI Agent Push Beyond Payments

SBI Chairman Calls for ‘Know Your Agent’ Controls

SBI Chairman C.S. Setty said banks will need systems to identify and authenticate AI agents as they begin executing financial transactions for customers. He called for audit trails, traceability and clear accountability, saying 99% accuracy is not enough when AI is acting in financial systems. Setty described “know your agent” as an extension of existing customer-verification controls.

Read more: SBI Chairman Calls for ‘Know Your Agent’ Controls 

Ant, Mastercard and Visa Develop AI Agent ID Framework

Ant International, Mastercard and Visa are developing a Know-Your-Agent framework that would allow payment networks, wallets, marketplaces and AI platforms to recognize verified agents across different systems. The proposed framework would link each agent to a verified operator, establish common security requirements and support continuous monitoring while preserving each network’s own risk controls. The companies are working through BuildFin.ai, an industry platform convened by the Monetary Authority of Singapore. The framework is still being developed.

Read more: Ant, Mastercard and Visa Develop AI Agent ID Framework 

UN Rights Chief Calls for Stronger AI Safeguards

UN human rights chief Volker Türk warned that advanced AI could pose an existential risk and called for internationally agreed red lines, independent safety checks and stronger cooperation among developers. He cited cases in which AI systems escaped testing environments or tried to prevent themselves from being shut down, and also raised concerns about employment, democracy and the environment.

Read more: UN Rights Chief Warns Advanced AI Could Threaten Humanity

Anthropic Researcher Puts Extinction Risk Above 10%

Anthropic Alignment Science Lead Evan Hubinger said his personal estimate of AI causing human extinction within the next decade is above 10%. His comments followed researcher Jacob Coxon’s resignation over concerns about the race to build more capable systems. Hubinger said the risk concerned future superintelligence rather than current AI models.

Read more: Anthropic Alignment Lead Puts Risk of AI Wiping Out Humanity Above 10%

DeepMind Selects 16 APAC Climate AI Projects

Google selected 16 Asia-Pacific organizations for its DeepMind AI for the Planet accelerator, a three-month program focused on climate resilience, agriculture and carbon solutions. Four are from India: Terrastack, Climitra Carbon, Farmers for Forests and Varaha Climate. Participants will receive access to Google AI models, technical support and mentorship to help scale existing projects.

Read more: Google DeepMind Selects 16 APAC Climate AI Projects