Mastercard's AI Shopping, AGNTCY Framework, and AI in Crypto Trading

Ever imagined a world where your shopping is completely handled by artificial intelligence? Well, that future might be closer than you think! Welcome to The AI Agent Daily Brief, your go-to for the latest AI updates. Today is Tuesday, April 29th, 2025. Here’s what you need to know about the next big leap in AI-driven shopping. Let’s dive in. Today, we're exploring Mastercard's latest innovation, Mastercard Agent Pay, which is set to revolutionize how we shop online.

The credit card giant has announced new tools designed to make it easier for artificial intelligence agents to handle card payments and make purchases on behalf of users. This is a significant step towards integrating AI into our everyday shopping experience. Mastercard is leveraging tokenized payment technology to ensure that your sensitive information, like credit card numbers, is protected.

This technology replaces your card details with secure tokens, similar to the system used by digital wallets like Apple Pay. It's all about keeping transactions safe while letting AI agents take the reins. The initiative comes as more companies are eager to find secure ways to let artificial intelligence handle purchases without introducing new risks.

Mastercard’s approach ensures AI agents are properly authorized to make purchases, sharing this verification with other payment companies and merchants. Partnering with tech behemoth Microsoft, Mastercard is developing more ways to expand AI shopping and is keen to collaborate with other AI companies soon. They're also teaming up with checkout providers like Braintree and Checkout.com, along with banks that issue credit and debit cards, to roll out these tools more widely.

As AI tools for online shopping grow, retailers are increasingly concerned about spotting fake or risky transactions. Mastercard's new tools aim to address these concerns, providing peace of mind for both consumers and businesses. On Wall Street, Mastercard is gaining attention as analysts have given MA stock a Strong Buy consensus rating. With the average price target suggesting a 17.1% upside potential, it seems investors are optimistic about the company's future in AI-driven commerce.

Imagine walking into a store and seeing price tags that seem to change with the blink of an eye. That's the power of Multi-Agent AI Pricing Systems, and it's shaking up retail like never before. This technology is rapidly becoming the go-to solution for retailers and e-commerce giants who want to stay ahead of the competition, protect their profit margins, and respond instantly to market changes. So, what exactly is a Multi-Agent AI Pricing System?

At its core, it's a network of autonomous artificial intelligence agents. Each agent is responsible for different pricing functions, working together—sometimes even competing—to make the best pricing decisions. These agents simulate various pricing strategies, analyze real-time data from competitors, and adapt to changes in customer demand almost instantaneously. It's a huge leap from traditional pricing software that might only update weekly or rely on static rules.

These AI agents are constantly learning, negotiating, and optimizing prices across thousands of stock keeping units, adjusting based on competitor pricing, customer buying behavior, time of day, and even seasonality. The result? A highly responsive pricing ecosystem that can operate at scale, providing the kind of agility that human pricing managers simply can't achieve alone. Who's already using this technology? Well, major players in the e-commerce and grocery sectors are leading the charge.

Amazon, for instance, has set the standard in dynamic pricing, believed to utilize multi-agent principles to automate price changes every 10 to 15 minutes across millions of listings. Walmart is also investing heavily in AI-driven pricing optimization through its tech lab and acquisitions like Jet.com. In Europe, companies like REWE Group in Germany and Carrefour in France are collaborating with AI startups to test predictive pricing systems using agent-based logic.

They're aiming to boost promotion effectiveness while cutting down on waste. Meanwhile, Ocado in the UK is exploring AI pricing integrated with its warehouse forecasting tools, especially in areas like fresh food where speed is crucial. Even smaller retailers and direct-to-consumer brands are getting on board, turning to software-as-a-service-based AI pricing platforms like Revionics, Quicklizard, and Pricemoov.

These platforms often use multi-agent frameworks behind the scenes, providing dynamic pricing at a fraction of the cost of building proprietary systems. Why are retailers so eager to adopt multi-agent AI? It's all about replicating real-life market dynamics. Imagine one AI agent focused on profit optimization, another on competitive parity, and a third on customer satisfaction.

The system weighs their input and makes a calculated pricing decision—at scale, across thousands of stores or e-commerce listings. Retailers gain speed, precision, resilience, and ultimately, profitability. But there are challenges, too. Implementing multi-agent AI pricing isn't cheap or simple. It requires cloud computing capabilities, integration with enterprise resource planning, point-of-sale, and customer relationship management systems, plus a robust data pipeline from multiple sources.

And let's not overlook the need for AI expertise, often sourced from external consultants or vendors. Depending on the scale, initial deployment costs can range from two hundred fifty thousand to two million pounds, especially for large retail chains. Smaller retailers using cloud-based platforms may pay ten thousand to fifty thousand pounds annually, depending on complexity and stock keeping unit volume. There's also the reputational risk.

Overly aggressive AI pricing could lead to price gouging, algorithmic collusion, or consumer backlash—issues that regulators and watchdogs are keeping a close eye on, especially in the United States and the European Union. So, is this the future of retail? In short, yes. As data becomes more abundant and AI more sophisticated, pricing will shift from being a monthly or daily task to a continuous, automated process.

Retailers who don't embrace this shift might find themselves outpriced and outpaced. However, the best-performing AI pricing systems aren't fully autonomous. They're augmented systems, designed to work alongside human merchandisers and category managers. As one pricing director at a leading French retailer put it, "The AI gives us the options. Humans make the final call. But the machine is learning faster than we ever could."

Multi-agent AI pricing isn't just an evolution—it's a revolution in retail pricing strategy. Retailers willing to invest in this technology stand to gain a competitive edge, operational efficiency, and deeper customer insight. The age of spreadsheet-driven pricing is over, and the question is not if retailers will adopt multi-agent AI pricing, but when—and how fast they can scale it to survive in an increasingly competitive global market.

AI agents are everywhere, performing countless functions, but they rarely talk to each other. That's where AGNTCY comes in, an open-source collective aiming to change the game. Imagine a world where AI agents can seamlessly connect and collaborate, both within and across organizations. AGNTCY is working to establish the protocols and standards for an interoperable "internet of agents" that could revolutionize how these digital workers interact.

Tushar Agrawal, senior director of artificial intelligence at Cisco, recently spoke at Fabrix.AI's Agentic Demo Day about this groundbreaking initiative. AGNTCY's collaboration includes major players like Cisco, LangChain, Galileo, Fabrix.AI, MongoDB, and Boomi. The goal? To create a network where AI agents can communicate effectively, much like microservices within a service-oriented architecture. So, why is agent connectivity so important?

Picture a team of agents collaborating to get tasks done. With AGNTCY's protocols, these agents could talk to each other, understand each other's capabilities, and interact securely with low latency. It's akin to the early days of the internet, where standardized protocols like Internet Protocol and Domain Name System enabled seamless communication across services.

Agrawal envisions an "internet of agents" that layers on top of the cloud internet, allowing cross-domain, cross-vertical, and cross-purpose collaboration. This architecture could transform how organizations deploy and manage AI agents, making it as effortless as setting up a website today. AGNTCY's framework includes three main components: Discover, Compose, and Deploy and Run.

The Discover component acts like a directory, helping agents find each other and sync across directories, similar to a DNS service. It utilizes an Open Agentic Schema Framework, standardizing how agents describe their capabilities for easy discovery and selection. Once agents are discovered, the Compose component helps bring them together, ensuring they operate correctly and securely.

This involves deploying agents and managing their runtime, ensuring they perform their intended functions without overstepping boundaries. Finally, the Deploy and Run component introduces protocols for secure, low-latency communication between agents. Using an agent gateway protocol or messaging protocol, this setup allows for a publisher/subscriber model, ensuring data is shared effectively and securely.

The lessons learned from the development of the internet provide a blueprint for aligning multiple AI agents to support business processes. Just as setting up a website is now second nature, AGNTCY envisions a future where deploying agentic applications is just as seamless, whether within an organization or globally.

Did you know that a staggering 96% of enterprises are planning to expand their use of artificial intelligence agents in the next year? It's true, and it's a clear indicator of how pivotal these technologies are becoming in the corporate world. Imagine a bustling office where performance optimization bots are streamlining operations, security monitoring agents are vigilantly protecting data, and development assistants are helping engineers code more efficiently.

That's not a scene from a sci-fi movie—it's the reality many companies are gearing up for. So, why does this matter? Well, according to Cloudera's latest report, most organizations believe that investing in AI agents is essential for maintaining a competitive edge. We're talking about a major shift from experimentation to delivering tangible business results.

Abhas Ricky, the chief strategy officer at Cloudera, put it perfectly: "AI agents have moved beyond experimentation—they’re now delivering real automation, efficiency, and business results." This isn't just about keeping up with the Joneses; it's about transforming operations with high-fidelity, well-managed data driving better outcomes. In 2025, agentic AI is taking center stage, building on the momentum of generative AI but with even greater operational impact.

Cloudera is at the forefront of this transformation, enabling global organizations to design secure, scalable, and integrated AI workflows that turn data into action. Let's zoom in on Singapore for a moment. A whopping 87% of business leaders there say their prior investments in generative AI have set the stage perfectly for implementing AI agents. They're using these agents for everything from fraud detection to patient monitoring and customer service. However, it's not all smooth sailing.

Many Singaporean enterprises are voicing concerns about AI bias and fairness. In fact, 71% are taking proactive steps by implementing safeguards like human review loops, diverse training data, and formal fairness audits to address these issues. With such a high percentage of enterprises expanding their use of AI agents, it's clear that this technology is not just a passing trend.

It's becoming an integral part of how businesses operate, offering both challenges and opportunities for innovation and growth.

Ever wondered if artificial intelligence could bring more transparency to crypto trading? Well, Oasis Protocol is doing just that. They've just rolled out WT3, a new kind of AI trading agent that promises to bring both privacy and verifiability to the crypto world. Crypto trading bots have been seen as mysterious black boxes, often raising concerns about how they operate. But Oasis Protocol’s WT3 is changing the game by operating inside a secure Trusted Execution Environment.

This means it can produce on-chain cryptographic proofs of its actions, allowing users to verify what the bot is doing without exposing its trading strategies. This is a significant step because it addresses the dual problem of privacy and verifiability that most crypto trading bots face. Typically, if bots run publicly, others can see all the trades, but if they remain private, users can't know for sure if the bot is doing its job correctly.

WT3 solves this by using Trusted Execution Environments to keep operations private while still producing verifiable proofs. Oasis Protocol’s WT3 is set to begin trading with an initial funding of one hundred thousand dollars from the Oasis Protocol Foundation. Users will soon be able to stake stablecoins to participate in strategy yields, with fifty percent of the returns supporting ROSE token buybacks and burns.

It’s a model that not only ensures privacy and transparency but also offers financial incentives for participants. The launch of WT3 marks a shift toward confidential, verifiable AI agents in decentralized finance, reflecting a broader trend of blockchain platforms embracing confidential computing. This means that the future versions of this bot will allow users to stake into strategies or even deploy their own customizable trading bots.

Matej Janež, Head of Partnerships at Oasis Protocol Foundation, highlighted the impact of these bots on the market. He pointed out that AI agents simplify the user experience in crypto trading, making it more accessible. This is particularly important as the crypto market has historically struggled with onboarding users due to complex interfaces and security concerns.

So what does this mean for the future of AI in crypto trading? As Oasis Protocol positions itself at the intersection of decentralized finance and AI, they’re setting a new standard for transparency and trust. It’s a move that could redefine how we think about AI in the crypto space, offering a solution to the transparency issues that have long plagued the industry. That’s it for today’s The AI Agent Daily Brief.

With Oasis Protocol unveiling verifiable AI agents for crypto trading, we're seeing a new era of transparency and trust in the crypto world. Thanks for tuning in—subscribe to stay updated. This is Michelle, signing off. Until next time.