Proposal: Measuring Offensive AI-Driven DNS Traffic
The goal of this hackathon project is to design and prototype an open-source measurement toolkit capable of detecting and characterizing DNS traffic patterns generated by AI-powered applications and automated agents. As AI systems increasingly perform automated web retrieval, API calls, and knowledge retrieval tasks, they may generate distinctive DNS query behaviors. These behaviors could include higher query volumes, novel query structures, or machine-generated patterns that differ from traditional human-driven browsing. However, the technical community currently lacks concrete measurement tools to quantify these impacts.
The proposed hackathon project will focus on three main objectives. First, participants will build lightweight instrumentation capable of analyzing DNS query logs to identify patterns potentially associated with AI-driven workloads. This may include detecting high-frequency automated queries, domain discovery patterns, or atypical query distributions. Second, the project will prototype classification techniques—using simple heuristics or machine learning—to distinguish between human-driven DNS activity and automated AI-generated traffic. Third, the resulting toolkit will produce aggregated metrics and visualizations that help researchers and operators evaluate whether AI workloads are introducing measurable changes to DNS traffic patterns.
The project will leverage publicly available DNS telemetry sources where possible and can be tested in controlled environments or on volunteered recursive resolver datasets. By the end of the hackathon, the team aims to deliver a proof-of-concept analysis pipeline, sample datasets, and a preliminary technical report summarizing observed patterns.
This work directly supports the broader Internet community’s need for evidence-based analysis of emerging trends. Rather than relying on speculation, the hackathon will provide an initial foundation for systematic measurement of AI’s interaction with DNS infrastructure. The resulting tools and findings could inform future discussions within the Internet Engineering Task Force and related technical communities, helping guide research, operational practices, and potential standards work to ensure that the DNS ecosystem remains stable, resilient, and prepared for the evolving AI landscape.
_______________________________________________Background: Researching the Impact of AI on the DNS
Why This Matters Now
During the joint meeting between the ICANN Board and the Root Server System Advisory Committee (RSSAC), held yesterday, March 10, 2026, at the ongoing ICANN85 meeting, a critical question was raised. Derived from ICANN’s Five-Year Strategic Plan, the Board requested that the RSSAC identify emerging trends capable of reshaping the Internet.
The most significant trend identified during this dialogue is the rapid rise of Artificial Intelligence (AI). While our primary focus is typically the "Root" , it is clear that AI will influence the entire Domain Name System (DNS) as established during the meeting. Based on the productive discussions held during that meeting, I would like to advance this conversation as follows:
Moving forward to Facts
Currently, there is much speculation regarding how AI will change the Internet; however, it is vital that we establish a foundation of technical fact. To ensure the Internet remains stable and secure, we cannot rely on assumptions. We require empirical research—the collection of real-world data and evidence. We need to analyze how AI-powered applications and automated systems are currently impacting the DNS so that we may adequately prepare for the future. Consequently, I propose the formation of a Working Group dedicated to conducting this essential research.
The Objectives/Goals of this Research
This proposal suggests a comprehensive study to examine the "big picture" of AI's interaction with the DNS. The objective is to produce a formal report that will, among other priorities:
- Track Traffic: Determine if AI is generating increased traffic volumes or unique types of requests and identify the genuine and the fake. To do this the RSOs may be engaged.
- Identify Early Risks: Detect whether AI integration is placing additional stress on system security or resilience with respect to DNS.
- Support Future Planning: Equip the global community of experts with the data necessary to develop informed, "fit-for-purpose" advisory document for the future.
By initiating this research now, we ensure that the Internet remains fast, reliable, and safe as AI technology continues to evolve.
This proposal is intended for discussion and refinement. The overarching objective is to generate articulated, empirical data on the impact of AI on the DNS, directly addressing the inquiry posed by the ICANN Board to the RSSAC during our meeting on March 9, 2026 at the ongoing ICANN85 Community Forum.
Thank you.
Olorundare, James Kunle
--
+2348036551591
rssac-caucus mailing list -- rssac-caucus@icann.org
To unsubscribe send an email to rssac-caucus-leave@icann.org
_______________________________________________
By submitting your personal data, you consent to the processing of your personal data for purposes of subscribing to this mailing list accordance with the ICANN Privacy Policy (https://www.icann.org/privacy/policy) and the website Terms of Service (https://www.icann.org/privacy/tos). You can visit the Mailman link above to change your membership status or configuration, including unsubscribing, setting digest-style delivery or disabling delivery altogether (e.g., for a vacation), and so on.