惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

Engineering at Meta
Engineering at Meta
Microsoft Azure Blog
Microsoft Azure Blog
I
InfoQ
OSCHINA 社区最新新闻
OSCHINA 社区最新新闻
人人都是产品经理
人人都是产品经理
让小产品的独立变现更简单 - ezindie.com
让小产品的独立变现更简单 - ezindie.com
T
Tailwind CSS Blog
MongoDB | Blog
MongoDB | Blog
Google DeepMind News
Google DeepMind News
WordPress大学
WordPress大学
量子位
美团技术团队
大猫的无限游戏
大猫的无限游戏
Cyber Security Advisories - MS-ISAC
Cyber Security Advisories - MS-ISAC
Last Week in AI
Last Week in AI
博客园 - 司徒正美
奇客Solidot–传递最新科技情报
奇客Solidot–传递最新科技情报
小众软件
小众软件
C
Check Point Blog
博客园 - 三生石上(FineUI控件)
N
Netflix TechBlog - Medium
Recent Announcements
Recent Announcements
有赞技术团队
有赞技术团队
月光博客
月光博客

Help Net Security

ChatGPT advanced account security adds passkeys and hardware keys Week in review: High-severity LPE vulnerability in the Linux kernel, cPanel 0-day exploited for months Automating Pentest Delivery: A Step-by-Step Guide - PlexTrac Open-source privacy proxy masks PII before prompts reach external AI services Shadow AI risks deepen as 31% of users get no employer training Identity is the control plane for distributed infrastructure AI traffic is getting bigger, louder, and less predictable New infosec products of the month: April 2026 cPanel zero-day exploited for months before patch release (CVE-2026-41940) Cisco releases open-source toolkit for verifying AI model lineage Met Police face criticism for using AI to spy on their own officers Nine-year-old Linux kernel flaw enables reliable local privilege escalation (CVE-2026-31431) Hacker with a special interest in breaching sports institutions ends behind bars - Help Net Security IP Fabric MCP server adds governance and control to enterprise AIOps workflows - Help Net Security Aqua Compass MCP server enables real-time investigation and containment of runtime threats - Help Net Security Google brings instant email verification to Android, no OTP needed - Help Net Security If cyber espionage via HDMI worries you, NCSC built a device to stop it - Help Net Security Apple fixes iPhone bug that let FBI retrieve deleted Signal messages(CVE-2026-28950) - Help Net Security GopherWhisper APT group hides command and control traffic in Slack and Discord - Help Net Security OpenAI tackles a bad habit people have when interacting with AI - Help Net Security A year in, Zoom's CISO reflects on balancing security and business - Help Net Security Scenario: Open-source framework for automated AI app red-teaming - Help Net Security GDPR works, but only where someone enforces it - Help Net Security Ransomware, fraud, and lawsuits drive cyber insurance claims to new peaks - Help Net Security Google’s Workspace Intelligence promises privacy while running on your data - Help Net Security Cyberattack on French government agency triggers phishing alert - Help Net Security Claude Mythos finds 271 Firefox flaws, Mozilla believes zero-days are numbered - Help Net Security Prove Identity Platform connects verification, authentication, and fraud prevention - Help Net Security New Mirai variants target routers and DVRs in parallel campaigns - Help Net Security Acronis GenAI Protection gives MSPs control over AI usage and data risks - Help Net Security
The assembly line behind 1.5 million malicious domains
Anamarija Pogorelec · 2026-06-12 · via Help Net Security

Attackers registered roughly 1.5 million malicious domains during the first five months of 2026. The registration patterns resemble industrial output. Most of the domains were created by attackers, put to use within weeks, and concentrated among a small set of registrars, top-level domains, and hosting providers.

malicious domain registration

New research examined more than 1.5 million unique domains flagged on VirusTotal between January and May 2026. Each domain was flagged by at least five independent VirusTotal scanning engines and first appeared on the platform during the study window. The detections were combined with WHOIS registration records, passive DNS resolution data, and the Tranco popularity ranking of well-known sites.

Close to nine in ten of the domains were registered by attackers for malicious use. The remainder were existing legitimate domains that attackers took over. Most of that smaller group qualified because of age, having been registered years before detection, a sign that long-established sites were repurposed for attacks.

Domains built for a short life

New malicious domains appeared at a steady, high volume each month. January 2026 recorded the most, with several hundred thousand domains, and the following months stayed in a similar range.

Attacker-created domains reached VirusTotal quickly after registration. The median domain was about two months old at first detection. A portion were detected within one day of registration, and close to a third within one week. This short window gives defenders limited time to identify a domain before attackers activate it.

Abuse concentrates at a few registrars and TLDs

A small number of registrars handled most attacker-created domains. The four busiest together covered more than a third of attack domains. The top ten registrars handled close to six in ten of the domains with known registrar data, with a long tail of several thousand additional registrars splitting the rest.

Top-level domain choices showed similar concentration. The .com extension led by a wide margin, accounting for roughly a third of all attack domains. It was followed by lower-cost generic and country-code extensions such as .top, .cc, and .xyz. The top ten extensions together covered about two-thirds of attack domains, with the remainder spread across hundreds of others.

This concentration points to intervention. Anti-abuse policies and faster takedown procedures at the leading registrars could reduce a large share of attacker-created domains.

The shared infrastructure problem

Most of the busiest hosting addresses belonged to Cloudflare. Eight of the top ten IP addresses hosting attack domains were Cloudflare addresses. The two busiest each hosted more than 230,000 distinct attack domains. These addresses are shared reverse-proxy endpoints used across Cloudflare’s network, so a single address serves many sites at once and hides the origin server behind it.

At the network level, Cloudflare’s autonomous system hosted the largest number of attack domains, with another large network and AWS close behind. Attackers gravitate toward reputable cloud and content-delivery providers because traffic from these networks is harder to block at the network level. The same services that protect legitimate sites from attack and surveillance also shield malicious ones, which makes takedown a question of provider cooperation.

DNS traffic concentrates in a small group

Query volume followed a steep distribution. Most attack domains drew modest traffic. A small group at the top drew enormous volumes, with the single busiest domain receiving more than two billion queries. This high-traffic group accounted for most of the query volume and the most user exposure. Sinkholing it would lower end-user risk substantially.

Registration in bulk

Coordinated mass registration appeared throughout the dataset. Grouping attack domains by registrar and creation date, any group of five or more domains sharing both attributes counted as a batch. More than three-quarters of attack domains with usable WHOIS records belonged to such a batch.

The largest single batch held more than two thousand domains registered with one registrar on a single day. Domain names within batches often followed short alphanumeric patterns, a sign of automated generation. Same-day registration of thousands of domains under one registrar points to scripts that produce and register names in bulk to assemble large attack fleets quickly.

Brand impersonation targets a few large names

Attackers embedded recognizable brand names into their domains. Brand tokens drawn from the most popular sites were checked against each attack domain name. WhatsApp was the most-copied brand by a wide margin, appearing in close to 20,000 attack domains. Google, Coinbase, and Bet365 also ranked among the leaders. The presence of a cryptocurrency exchange and a gambling platform points to credential harvesting and phishing aimed at financial accounts.

The matching method used simple substring detection, so some entries reflect common letter sequences more than deliberate impersonation. The distinctive brand names in the results, led by WhatsApp, carry more weight. A small share of all attack domains contained a recognizable brand token, spread across thousands of distinct brands.

Where defenders can intervene

The good news is that the same concentration that makes this work efficient for attackers also gives defenders a short list of places to push. If a handful of registrars are issuing most of the domains, rate limits and anti-abuse checks on same-day bulk registrations would slow a lot of the activity at once. The patterns are easy to spot: thousands of names with the same registrar and the same creation date, often with machine-generated spellings.

The hosting side calls for a different kind of work. Cloudflare and AWS both run abuse-reporting programs, and the volume here suggests those programs need automated pipelines that connect threat intelligence teams directly to the providers, so flagged domains tied to known malicious hosting get pulled faster. The busiest domains, the small group pulling billions of queries, are the ones worth sinkholing first, since they reach the most people. And the brands attackers lean on most, WhatsApp and Google among them, are obvious candidates for automated monitoring.

The throughline is that domain abuse runs like a production line with a few chokepoints. A few registrars issue most of the domains. A few extensions host most of the names. A few networks carry most of the traffic. Press on those points and you reach a large share of the problem. The researchers released their annotated dataset publicly for anyone who wants to dig further.

Apply today: Simplify security management with CIS SecureSuite Platform