February 2026 Summaries

Businesses are increasingly encountering a challenge known as the agentic commerce problem, where AI traffic comprises both beneficial and potentially harmful agents. Useful AI agents assist customers and enhance business operations, while others may engage in malicious activities such as content theft or fraud. A significant portion of AI agents, 80% according to DataDome's threat research team Galileo, fail to properly identify themselves, complicating the detection of legitimate versus harmful interactions. To address this, businesses are encouraged to adopt bot and agent trust management software, which provides real-time visibility, control, and classification of AI agent interactions. By implementing such platforms, companies can manage agent traffic through various capabilities, including agent verification, intent-based detection, Model Context Protocol (MCP) protection, and dynamic access controls. Additionally, monetization of high-value content through differential pricing for AI agent access can present new revenue opportunities. DataDome offers a comprehensive solution that integrates these features, aiming to balance support for legitimate agents with protection against potential abuses.

AI agents are transforming the online shopping landscape by acting as intermediaries for consumers, conducting product research, price comparisons, and completing purchases, which has led to a significant trust paradox for merchants. While these AI agents, such as ChatGPT, can enhance the shopping experience, they also open the door for malicious activities like data scraping and fraud, as they often mimic legitimate automated traffic. This has necessitated the development of authentication protocols like Web Bot Auth, which use cryptographic signatures to verify agent identities, and commerce-specific extensions such as Visa's Trusted Agent Protocol and Mastercard's Agent Pay, which ensure that agents are authorized to transact on behalf of specific consumers. However, identity verification alone is insufficient, as authenticated agents can still engage in harmful behaviors, making behavioral intelligence critical to distinguish between legitimate and malicious activities. DataDome addresses this by providing real-time agent classification and trust management, allowing merchants to balance security with the need to maintain sales, thus protecting against fraud while enabling legitimate agent-driven commerce.