Woofun AI reports that the emergence of autonomous commerce has exposed a critical regulatory vacuum, prompting the deployment of blockchain-based 'internet courts' to adjudicate disputes between AI agents before they escalate into systemic failures. The proliferation of intelligent agents capable of executing complex economic tasks—from Robinhood’s autonomous stock trading and SAP’s Joule managing corporate procurement to Amazon’s 'Buy for Me' negotiating delivery terms—has accelerated the need for automated conflict resolution. While major technology and cryptocurrency entities including Anthropic, OpenAI, Coinbase, and Circle are actively building the infrastructure for this agent-driven future, the underlying legal framework remains fragmented. When an agent-ordered sofa arrives damaged or in the wrong color, or when a seller disputes liability for post-delivery issues, the resulting friction threatens to stall the entire ecosystem. Although software can now represent humans in shopping, hiring, and payment execution, the inherent risk of AI 'hallucinations' and the non-monetary complexities of business interactions create scenarios where traditional human-centric legal recourse is either too slow or too expensive. This structural mismatch between the speed of autonomous transactions and the latency of human judicial processes has positioned decentralized dispute resolution as a necessary component of the emerging digital economy. The core challenge lies not in the ability of agents to transact, but in the absence of a standardized, scalable mechanism to enforce contracts and resolve disagreements when those transactions fail to meet predefined expectations.

The conceptual foundation for this solution was articulated by David Riudor, CEO and co-founder of the GenLayer Foundation, which operates from the Cayman Islands. Riudor identified that agent-based commerce is at a critical turning point, yet the industry lacks preparedness for the potential outcomes of autonomous interactions. To address this, the foundation launched GenLayer, a blockchain network featuring a core application known as the internet court. This system is designed specifically to resolve disputes between agents without human intervention, leveraging a decentralized network of validators to ensure impartiality. The initiative has garnered support from 26 prominent crypto and AI companies, including the exchange OKX, wallet provider MetaMask, and Binance’s BNB chain, signaling broad industry confidence in the viability of automated adjudication. The internet court does not aim to replace human judges entirely but rather to provide a specialized layer for high-frequency, low-value disputes where traditional legal engagement is economically unfeasible. By enabling agents to sign contracts with clear, pre-defined terms and automatically triggering an AI jury evaluation when agreements break down, the system offers a rapid alternative to litigation. This approach ensures that minor conflicts do not accumulate into significant operational bottlenecks, allowing the broader ecosystem of autonomous commerce to function with greater reliability and efficiency.

The economic rationale for such a system is underscored by the stark disparity between the cost of traditional legal services and the value of individual agent transactions. Albert Castellana, co-founder and CEO of GenLayer Labs, emphasized that the platform is not competing with traditional legal systems but offering a complementary alternative for cases where hiring a lawyer is not economical. For a claim valued at $10,000, the cost of legal representation often outweighs the potential recovery, yet ignoring the issue results in losses. In contrast, the internet court can resolve such disputes quickly, potentially costing just a few cents. This cost efficiency is critical given the explosive growth in AI-driven commerce. Since October 2024, retail website traffic driven by AI recommendations has increased by more than 14 times, indicating a rapid shift in consumer behavior. McKinsey predicts that by 2030, AI agents could facilitate between $3 trillion and $50 trillion in global consumer transactions.

However, most existing infrastructure focuses on the 'smooth path'—where agents successfully find products, complete payments, and receive goods without incident. The internet court addresses the 'friction path,' providing a mechanism to handle the inevitable errors and disagreements that arise in such a high-volume environment. By reducing the cost of dispute resolution to near zero, the system enables agents to engage in more complex and risky transactions without fear of catastrophic loss from unresolved conflicts.

Currently, the internet court has been applied in limited but high-stakes scenarios, particularly in verifying content authenticity. The social platform Collective Memory uses the GenLayer system to reward users for uploading real-time photos, videos, and reports, while also addressing disputes over the authenticity of submitted media. For instance, when determining whether a disputed image is fake, the system evaluates factors such as upload time, location, related submission records, and the user’s historical activity. This capability is crucial for verifying content from conflict zones, such as videos from bombed schools in Gaza or scenes from streets in Tehran, where misinformation can have severe real-world consequences. By providing an objective, algorithmic assessment of evidence, the internet court helps maintain the integrity of user-generated content platforms. This application demonstrates the system's versatility beyond simple commercial disputes, extending to areas where trust and verification are paramount. The ability to automatically assess the credibility of digital evidence sets a precedent for how AI-driven platforms can manage content moderation and dispute resolution at scale, reducing the burden on human moderators and increasing the speed of verification.

The operational workflow of the internet court is built on a sophisticated jury mechanism that leverages blockchain technology to ensure fairness and transparency. At the heart of this system is a panel of 5 randomly selected blockchain validators, each running a different AI model, such as Claude, GPT, or Gemini. One validator acts as the leader, proposing an initial decision, while the others vote secretly before publicly stating whether they agree. If a consensus is reached, a 30-minute dispute window begins, during which agents or humans can file challenges by paying a bond. If challenged, the jury size increases to 11 members, and the process repeats until a consensus is formed and no further objections are raised. This mechanism draws on the jury theorem proposed in 1785 by the Enlightenment philosopher and mathematician Nicolas de Condorcet, which posits that the more independent evaluators there are, the higher the probability of reaching a correct conclusion. By combining multiple AI models, GenLayer believes the system is harder to manipulate than a single model or a single human arbitrator. This layered approach to decision-making ensures that errors or biases in one model are offset by the diversity of the others, enhancing the overall reliability of the rulings.

Woofun AI data shows that despite its futuristic premise, the internet court is already online and in the testing phase, processing significant volumes of transactions. According to Castellana, the network handles around 350,000 transactions per day, resulting in 20,000 to 25,000 decisions. The platform is planned to be officially launched later this year, with tokens issued to attract more validators, allowing anyone to participate in the adjudication process. Beyond agent-based commerce, Riudor envisions the system expanding into prediction markets, where speed and accuracy are critical. For example, Polymarket currently relies on the UMA protocol to let token holders vote on dispute outcomes, but AI-assisted rulings would be faster and more scalable. Castellana noted that they are in talks with several large prediction markets, which are waiting for the platform to go fully live but are already evaluating its potential. This expansion highlights the broader applicability of automated dispute resolution in decentralized finance, where traditional legal mechanisms are often too slow or jurisdictionally complex to be effective.

However, experts caution that the reliance on AI for adjudication introduces new risks, particularly regarding hallucinations and data bias. Andrew Hall, a professor at Stanford Graduate School of Business and research advisor to Andreessen Horowitz’s crypto team, wrote earlier this year that while large language models cannot be bribed and their performance is improving rapidly, they can still suffer from hallucinations. He warned that these models may be manipulated through clever prompts or contaminated training data, potentially leading to incorrect rulings. Lindsay Lin, former legal counsel at New York-based crypto venture capital firm Dragonfly and current CEO, echoed these concerns, noting that many large language models are interconnected due to shared training data and common flaws. This interconnectivity reduces the independence of AI jurors compared to human judges, who tend to have more diverse perspectives. Nevertheless, Lin acknowledged that people prefer using AI to handle low-value disputes because it is cheaper and faster than human jurors. Given the massive scale of agent-based commerce, the volume of disputes will be high, making standardized protocols for agents essential to clarify terms of cooperation and remedies when transactions are not completed properly.

The industry is responding to these challenges by developing standardized protocols to govern agent interactions. Just two weeks ago, the International Centre for Dispute Resolution of the American Arbitration Association announced the introduction of a 'legal context protocol' standard for agents. Developed in collaboration with Denver-based blockchain company Integra Ledger, the initiative includes founding contributors such as Google, IBM, Circle, and Ava Labs. These standards aim to provide a common framework for agents to understand legal obligations and dispute resolution processes, reducing ambiguity and potential conflicts. The widespread adoption of these protocols will depend on the ability of AI models to effectively reduce hallucinations and biases, as well as on the willingness of platforms to integrate them into their systems. By establishing clear rules for agent behavior, the industry can mitigate some of the risks associated with autonomous commerce and create a more predictable environment for both businesses and consumers.

Meanwhile, the infrastructure for agents to find, hire, and pay each other is rapidly taking shape, further accelerating the need for robust dispute resolution mechanisms. In recent weeks, GenLayer partner OKX and the AI-focused Near Protocol blockchain team have launched markets that allow agents to hire other agents to carry out paid tasks, such as obtaining datasets or assisting with code reviews. These markets enable a new form of labor exchange, where AI agents can collaborate and compete to complete tasks, creating a dynamic ecosystem of autonomous services. As this ecosystem grows, the frequency and complexity of disputes are likely to increase, necessitating more sophisticated and scalable resolution tools. The internet court’s ability to handle these disputes efficiently will be critical to the long-term viability of agent-based commerce, ensuring that the benefits of automation are not undermined by legal and operational friction.

Real-world legal precedents are also beginning to emerge, highlighting the regulatory challenges facing agent-based commerce. In November 2025, Amazon sued Perplexity, accusing its AI-powered Comet browser of logging into customer accounts, disguising itself as a regular Google Chrome browser, and making unauthorized purchases on Amazon’s platform. In March, a federal judge in California issued a preliminary injunction banning Comet from shopping on Amazon, but an appeals court later suspended the order while reviewing Perplexity’s appeal. This case underscores the difficulty of regulating AI agents that act on behalf of users across platforms, particularly when there is no unified enforcement mechanism. Regardless of the final court decision, the incident illustrates the broader challenge of defining liability and accountability in an era of autonomous agents. As millions of AI agents begin to act on behalf of users, the lack of clear legal frameworks could lead to widespread confusion and conflict, making the development of standardized, automated dispute resolution systems like the internet court increasingly urgent.