DataDome vs FouAnalytics

Introducing DataDome’s AI Key Takeaways Feature: Understand Your Protection at a Glance

DataDome has introduced a new AI Key Takeaways feature that provides users with a quick, at-a-glance understanding of their protection status. This product update is highlighted on the company's blog.

DataDome Joins Experian Agent Trust Ecosystem

Experian, a global data and technology company, announced that DataDome, a leader in bot and agent trust management, has joined its Agent Trust™ ecosystem. This collaboration aims to strengthen trust in agentic commerce by combining Experian's trusted identity foundation with DataDome's continuous intent verification. The partnership creates a layered trust framework that helps businesses verify AI agents' identities and ensure their behavior remains consistent with delegated authority. DataDome integrates directly with Experian's framework, using verified Know Your Agent signals to dynamically assess agent behavior. This allows organizations to make trusted decisions throughout the transaction lifecycle without re-architecting their technology stacks. Together, they aim to enable safe and scalable agentic commerce by stopping malicious or compromised agents in real-time while allowing legitimate transactions to proceed.

• DataDome joins the Experian Agent Trust ecosystem.
• The collaboration combines Experian's identity foundation with DataDome's continuous intent verification.
• DataDome integrates with existing commerce and security infrastructure without requiring re-architecture.

Designing Products for Headless AI Agents

The article discusses the shift from chat-based AI interfaces to headless AI agents that interact with products via APIs without a graphical user interface. It emphasizes the need for service design, API literacy, and careful contract design to accommodate these invisible users. Citing reports from DataDome, Cloudflare, and Fastly, it highlights the rapid growth of AI agent traffic and the importance of designing for failure states, idempotency, and semantic clarity. The author argues that product design must expand beyond screens to include the data contracts and error messages that agents encounter, urging teams to understand agent behavior through traffic analysis and sequence mapping.

• DataDome's AI Traffic Report Q2 2026 counted 17.7 billion AI agent requests in April, May, and June, up 45% from the previous quarter.
• Cloudflare's one-year report found that as of June 2026, 52% of crawler requests were for AI training, up from 22% in spring 2025.
• Fastly reported in June 2026 that AI traffic grew approximately 6.5 times faster than human traffic on its network between January and May.

FouAnalytics and fraud0 Partner for Automated Google Ads Refunds

FouAnalytics, an independent analytics platform for digital ads, announced a deeper partnership with fraud0 to bring an AI agent for automated Google Ads refunds to large advertisers. The AI agent audits Google Ads spend for invalid traffic, false clicks, and billing errors, automatically filing refund claims. fraud0 has audited over $2 billion in digital ad spend and recovered significant losses for clients like Uriach. The partnership aims to provide transparency in Google Ads, which previously lacked third-party measurement. Advertisers can activate the solution by adding a JavaScript tag and connecting their Google Ads account.

• FouAnalytics and fraud0 partnered to offer automated Google Ads refunds.
• The AI agent is the first to automate refunds for invalid clicks and billing errors.
• fraud0 has audited $2 billion+ in digital ad spend.

FouAnalytics Exposes Online Video Ad Completion Fraud

Dr. Augustine Fou's column on State of Streaming presents forensic data from FouAnalytics revealing widespread fraud in online video (OLV) advertising completion metrics. The analysis shows that users often leave pages within 0.5 to 17 seconds, making 30-second ad completions impossible, yet platforms report 90%+ completion rates. Fraudsters exploit a loophole by invoking complete.gif and other tracking pixels without actual ad view. Data indicates only 1-6% of video ads reach thirdQuartile, and nearly 0% are audible. Fou recommends advertisers demand independent verification tags or buy from reputable publishers like Disney and Netflix, and consider shifting budgets to display ads for cost-effectiveness.

• Users leave pages within 0.5 to 17 seconds, making 30-second video ad completions impossible.
• Fraudsters falsify completion rates by invoking complete.gif and other tracking pixels without actual ad views.
• FouAnalytics data shows only 1-6% of video ads reach the thirdQuartile mark.

FouAnalytics Exposes Rampant CTV Ad Fraud on Mobile, Web

Dr. Augustine Fou's FouAnalytics column reveals widespread Connected TV (CTV) ad fraud, where advertisers paying for premium CTV inventory receive ads on websites, mobile apps, wireless IP addresses, or even completely fake bid requests. The analysis shows that cheap CTV CPMs often hide mis-marked inventory, audience extension arbitrage, and fabricated delivery. Fou emphasizes the unreliability of platform and legacy verification vendor reporting, advocating for independent forensic measurement to ensure buys reach actual living-room screens. The piece provides concrete examples of campaigns with mobile app delivery, wireless ISP sources, missing household IPs, and undetectable sources, urging buyers to audit their campaigns and demand refunds or corrections.

• FouAnalytics found CTV impressions delivered to websites and mobile apps despite advertisers paying CTV prices.
• 5–25% of CTV impressions run on wireless IP addresses instead of household IPs.
• 61% of CTV impressions in one campaign had no detectable source.