AI agents

Consumers increasingly trust AI agents with shopping tasks, with 74% preferring them over friends for purchasing decisions. AI agents can negotiate, resolve issues, and manage subscriptions within set permissions. While delegation is rising for routine tasks, full autonomy in payments remains low. Consumers prioritize data safeguards, clear permissions, and recourse options. Generative AI is expected to significantly influence spending, with consumers seeking AI agents that help achieve an “idealized self.” Physical stores will evolve, focusing more on engaging experiences. Overall, consumers are selectively delegating, retaining control over high-risk or personally significant purchases.

Coinbase has launched “Coinbase for Agents,” enabling AI models like ChatGPT to execute cryptocurrency trades and autonomous payments using natural language. This tool leverages Coinbase’s x402 protocol for machine-to-machine payments, facilitating tasks like accessing research and compute resources without human intervention. The move aims to position AI agents as future economic drivers and financial actors, with Coinbase benefiting from trading fees and payment processing.

OpenAI has acquired Ona, a startup specializing in secure AI agent environments. This move enhances OpenAI’s Codex coding assistant, enabling it to handle more complex tasks and streamline agent deployment for organizations. The acquisition bolsters OpenAI’s aggressive growth strategy ahead of its anticipated IPO, signaling a focus on integrating diverse technologies to solidify its market position.

Visa’s integration with ChatGPT enables AI agents to autonomously execute financial transactions, removing human intervention from retail sales funnels. This innovation bypasses traditional closed-loop systems, connecting LLM reasoning directly to a universal transaction network. Retailers must adapt by structuring data for machine readability and optimizing for AI, as personalization shifts to AI profiles. Visa’s tokenization ensures secure, automated checkouts, while headless commerce facilitates millisecond transactions. New analytics are needed to track AI queries, and loyalty programs must be embedded into AI workflows for effective customer retention.

Shell is transitioning to fully automated predictive maintenance with C3 AI’s advanced agents, moving beyond basic anomaly detection. This strategic move aims to enhance operational efficiency, reduce downtime, and unlock significant economic value across its critical assets. The AI agents will autonomously manage the maintenance lifecycle, from identifying anomalies to completing repairs, minimizing human oversight and optimizing resource allocation. This advancement promises reduced unplanned downtime, improved safety, and extended equipment lifespan.

Morgan Stanley is opening its wealth management platforms to thousands of corporate clients’ AI agents, allowing direct data extraction from stock administration systems. This pioneering move bypasses traditional interfaces and signals a shift towards agentic AI in financial services. By next year, all 3,400 administration clients will have this capability, positioning Morgan Stanley as a leader in integrating AI for client and internal operations.

Meta is expanding its AI focus beyond advertising to target the AI agent market with Meta Business Agent. Integrated into WhatsApp, Messenger, and Instagram, it will automate customer interactions for businesses of all sizes, from inquiries to bookings. This move diversifies Meta’s revenue, aligning with CEO Mark Zuckerberg’s vision to compete with AI leaders and offer advanced business management capabilities through evolving agentic technologies.

Perplexity CEO Aravind Srinivas asserts that AI success hinges on maximizing economic value from computational power, measured by “token value per watt per user.” This involves balancing accuracy, latency, cost, privacy, and intelligence. Perplexity’s strategy focuses on “agentic AI” and an “orchestrator” tool that intelligently routes processing to optimize efficiency, aiming for on-device computation and a platform-agnostic approach to build a durable advantage in the competitive AI landscape.

Google Pay is evolving its payment infrastructure with the Universal Commerce Protocol (UCP) and a new server architecture. This shift aims to support transactions initiated by AI agents, moving beyond human-centric checkout processes. The updates include standardized machine-to-machine communication, a central merchant platform for data insights, enhanced Android callbacks, and expanded WebView support. This redefines customer journeys, demanding machine-readable product information and introducing cross-device biometric authentication for AI-driven transaction security.

Alibaba is advancing its AI ecosystem with the new Zhenwu M890 chip, designed for AI agents requiring extensive memory and inter-model communication. This hardware development is part of a multi-year roadmap and is complemented by an advanced large language model, Qwen 3.7-Max. This integrated strategy aims for self-sufficiency in AI infrastructure, reflecting a shift from procurement to long-term capability building in semiconductor development.