Font size:
Page background:
Letter spacing:
Images:
Disable visually impaired version close
Version for visually impaired people
News

Humans must verify AI solutions and determine ethical rules

The 89th conference of the Asia Pacific Top-Level Domain Association (APTLD) kicked off in the capital of Jordan, Amman, on February 2, 2026. APTLD brings together ccTLD registries from countries in the Asia-Pacific region. Representatives of the Coordination Center for TLD .RU/.РФ are also attending APTLD89.

Deputy Director of the Coordination Center Irina Danelia spoke during the session titled How ccTLD Managers Should Think and Plan for Business Continuity. She shared the Russian registry’s experience in these matters and how the Coordination Center develops, updates and puts to test its business continuity plans. Irina Danelia noted that testing all plan components is imperative, primarily to reveal and eliminate flaws but it is also useful for employees’ practical training in on-the-spot action.

Olga Baskakova, Project Manager of the Coordination Center and Director of Domain Patrol, participated in a session called Evidential Thresholds – Action and Automation. She spoke about the Coordination Center’s experience with countering DNS abuse in .RU and .РФ, in particular by differentiating threats and balancing between prompt response and prevention of erroneous decisions. She noted that in the Russian domain ecosystem, there is a unique interaction experience between trusted experts and registrars. Every expert selects their own set of methods, tools and algorithms. A registrar has the right to cease delegation but is not obliged to do it. As a result, the minimum scope of abuse evidence will be determined by the registrar/expert duo and will depend on the level of trust.

Olga also highlighted that Domain Patrol experts currently use automation where machine learning may prove useful, in tasks such as log analysis, traffic monitoring, etc. Artificial intelligence is very often used for Big Data processing, the most important AI applications being behavioral model analysis and model forecasting and development in the context of potential threats. However, the role of humans remains extremely important:

Humans must determine the ethics and always verify AI solutions. Context analysis is very important in DNS abuse fallout mitigation, and the human brain is irreplaceable here,” Olga concluded.

APTLD89 continues – Follow our updates!

Previous News Next news