In Depth: AI Agents Ignite Global Battle for the Future of Payments

1. The article opens with examples highlighting rapid advancements in AI-driven commerce: Alibaba’s Vice President Wu Jia used the Qwen AI app to order and pay for milk tea with a voice command, and a Google executive in the U.S. used Gemini for birthday party planning and in-app purchases. These staged demonstrations underscore a shift toward AI agents handling real-time transactions entirely within chat interfaces, setting the stage for a new era where commerce and payments are increasingly agent-mediated. [para. 1][para. 2][para. 3]

2. This development signals a budding transformation in digital commerce, as the mechanism shifts from human-initiated payments to agent-to-agent (A2A) transactions. Globally, both technology and financial firms are racing to seize control over these AI gateways that could soon dominate global commerce. Events such as the U.S. Federal Reserve’s 2025 Payments Innovation Conference focusing on AI, Mastercard’s Agent Pay rollout, and innovations by Turkey’s Garanti BBVA exemplify this momentum. [para. 4][para. 5]

3. Consumer adoption is also surging. A 2025 Visa survey found that 74% of Asia-Pacific consumers already use AI for shopping decisions—a figure that rises to 86% in China and is predicted to reach 94%. As AIs become decision-makers, companies face urgent challenges: defining A2A rules and addressing related dilemmas in security, fraud, and regulation, while the structure of global commerce itself is reshaped. [para. 6][para. 7]

4. The current race mirrors the past contest to standardize QR code payments, but the new battleground is building protocols for AI agents to negotiate and execute payments. Tech giants like Google, Alibaba, and OpenAI are positioning their AI systems as the principal gateways to commerce. The advantage gained here means access to user traffic, commercial data, and revenues. [para. 8][para. 9]

5. Alibaba exemplifies aggressive adaptation, integrating its Taobao and Alipay platforms into the Qwen AI, thus creating a comprehensive AI-powered transactional hub. Its “AI Pay” feature achieved over 100 million users and processed 200 million transactions during the 2026 Chinese New Year. Competitors including JD.com, China UnionPay, as well as global names like OpenAI, Stripe, Mastercard, and Visa, are developing similar offerings. Industry experts note that having no AI capabilities in payments will render companies obsolete, as payment initiation will soon shift to AI orchestrators. [para. 10][para. 11][para. 12][para. 13][para. 14]

6. To enable seamless, cross-platform AI commerce, the industry is engaging in a standards war reminiscent of the QR code era. Google leads a consortium developing universal protocols for commerce and agent communication, supported by more than 100 partners including MasterCard and PayPal. OpenAI and Stripe are advancing a secure, token-based protocol for ChatGPT, while Ant Group’s Agentic Commerce Trust Protocol powers Alipay’s AI Pay. The fundamental challenge remains building trust, verifying identity, and ensuring compliance as non-human agents take center stage. [para. 17][para. 18][para. 19][para. 20][para. 21][para. 22]

7. Despite hype, truly autonomous AI payments are still emerging. Most current systems are agent-assisted rather than full A2A operations. For instance, Alipay’s AI Pay requires user confirmation, and Qwen’s system primarily integrates existing infrastructure with AI at the interface. The broader financial ecosystem is still designed for humans, so the prevailing model is “agent-to-human” interaction, not full agent autonomy. Full A2A transactions are described as an “unexplored wilderness” by experts. [para. 23][para. 24][para. 25][para. 26][para. 27][para. 28]

8. Risks associated with AI payments are surfacing. In one incident, an OpenAI engineer’s trading agent mistakenly transferred $450,000 in crypto to a scammer after losing session memory. Such vulnerabilities, including hallucinations, fraud, and security breaches, are escalating as AI grows more autonomous. Experts warn of scalable fraud via deepfakes and synthetic digital personalities. A regulatory paradigm shift is suggested: “Know Your Agent” might replace “Know Your Customer,” and blockchain could provide traceability. Until new frameworks mature, human authentication and compliance are essential safety mechanisms, and establishing clear liability will be critical before full AI-driven payments become mainstream. [para. 29][para. 30][para. 31][para. 32][para. 33][para. 34]