The most complete, continuously-updated database of AI-tool domains on the web. Classified into 18 functional categories. Built from 102 million domains. Updated every day.
Two years ago, blocking AI tools meant adding five domains to a firewall rule. That era is over.
The AI-tool landscape now numbers in the tens of thousands: specialized writing assistants, code generators, voice cloners, data-extraction bots, autonomous agents, and tools so niche they serve a single vertical. New ones register every day.
An employee pastes a confidential contract into a translation tool you've never heard of. A developer routes source code through a debugging service that launched last week. The data leaves before anyone knows to look.
GDPR, CCPA, DORA, and industry-specific data-residency rules hold organizations accountable for where data goes — not whether IT knew about the destination. Shadow AI is shadow IT accelerated: free, no installation, browser-tab access.
Static blocklists can't solve this. They're authored reactively, cover the famous names, and miss the long tail. The long tail is where nobody is looking — and where the risk lives.
We maintain a proprietary database of 102 million classified domains. Every domain is classified by function, content type, and risk category. As part of keeping this database current, approximately 300,000 newly registered domains are ingested and classified every day.
Any domain in this corpus — existing or newly added — that is classified as an AI tool automatically enters the AI-tools feed. This is how we catch AI tools that exist as domains but appear in no directory, no press coverage, and no curated list.
Separately from the 102M database, we run a dedicated pipeline that monitors sources specifically frequented by AI tools: app directories, product-launch platforms, developer communities, academic repositories, and the open web. Overlapping sources by design.
This pipeline finds AI tools that may not yet have enough web presence for the corpus classifiers to catch — tools announced yesterday, in beta, or serving a niche vertical. They're identified, classified, and added to the feed within the same daily cycle.
No manual list can match this cadence. The question isn't whether a curated list misses tools — it's how many, and for how long.
The 102M-domain corpus catches what exists. The discovery pipeline catches what's new. Both run every day.
102 million domains already classified by function, content, and risk.
~300K new domains ingested daily from zone files, CT logs, and registration feeds to keep the database current.
Any domain — existing or new — classified as an AI tool is automatically extracted into the feed.
Classified AI-tool domains sent to the shared delivery step below.
App directories, product-launch platforms, developer communities, academic repos, open web — scanned daily.
Candidate domains classified into 18 categories with tool name and subcategory assignment.
Dead domains pruned. Aliases and subdomains resolved. Clean, deduplicated set.
New AI-tool domains merged into the shared delivery step below.
Both pipelines merge here. One unified, deduplicated feed. CSV, Parquet, or API. Daily or weekly refresh.
Structured fields designed for direct integration into web filters, endpoint agents, and governance platforms.
Simple ingestion into any system
High-performance analytics pipelines
Real-time queries with category & date filters
We classify the entire AI-tool landscape into 18 functional categories and 140+ subcategories, so you can block or monitor by capability — not just by name. Click any category to see its subcategories.
View Full Taxonomy ReferenceDrop the AI-tools list into your categorization engine as a specialized feed. Your customers get AI-tool blocking out of the box. Differentiate with coverage that extends into the long tail.
Push the domain list to managed devices for browser-level or DNS-level blocking. Ideal for corporate hardware, BYOD, and school-managed tablets.
Build enforceable AI-use policies with a list that's actually complete. Know which tools employees access, which categories are in use, and which new tools appeared this week.
Block AI chatbots and essay generators while allowing educational tutors. Our 18-category taxonomy lets you set per-category policies: block Text & Language, allow Education & Learning.
Feed AI-tool domains into your data-loss prevention or CASB to flag data exfiltration to AI services. Catch sensitive data flowing to tools traditional threat intel doesn't cover.
Map AI-tool usage to regulatory requirements. Structured evidence: which tools are accessed, what they do, when they appeared. Audit trails, not anecdotes.
Pricing depends on license type (internal use vs. OEM/embedded vs. redistribution), deployment scope, and refresh cadence.
Get the AI-tools domain list as an independent feed. Ideal if you already have web categorization and need specialized AI-tool coverage.
The AI-tools database bundled with our complete 102M-domain categorization database. For general web categorization — not just AI tools.
Tell us your use case and we'll quote within one business day. Want to see the data first? Download the sample CSV.
Not scraped from a directory. Not compiled from press coverage. Built through systematic pipeline designed to find tools nobody's heard of yet.
We monitor the full landscape — domain registrations, app directories, product-launch platforms, developer communities, academic repositories, and the web itself. Overlapping sources by design.
Resolve every entry to its root domain, collapse aliases, and deduplicate. One clean entry per tool — not inflated counts from subdomains and mirrors.
Tools die. Domains park. Startups pivot. We continuously recheck every domain. The count represents live, operational tools — not a cumulative count including the dead.
Each domain passes through the same classification pipeline powering our 102M corpus. Assigned a functional category based on what it actually does, not what it claims — and flagged as AI-native or AI-enabled so you can separate purpose-built AI tools from platforms with embedded AI.
Cross-reference against the full 102M corpus. Catches AI tools that exist as live domains but appear in no directory — the completeness step directory-only approaches miss.
The pipeline runs every day. New tools enter. Dead tools leave. 300,000 newly registered domains swept into this cycle daily. A living asset, not a static snapshot.
200 well-known AI-native tools exactly as they appear in the delivered CSV — root domain, primary category, status, language, AI type, and every category > subcategory assignment. The full database additionally flags mainstream platforms with embedded AI (ai_enabled) separately from AI-native products, so you can block precisely or monitor broadly.
| root_domain | primary_category | active | lang | ai_type | categories |
|---|---|---|---|---|---|
| Loading sample data… | |||||
Anyone can compile a list of well-known AI tools. The question is what happens with the other 15,001. Our advantage is structural: we already classify 102 million domains. The Enterprise AI Tools Database is a specialized view of a corpus we maintain anyway. Every tool that exists as a domain passes through our classifiers — whether or not anyone has written about it.
Completeness and freshness aren't features we added. They're structural properties of how the data is built.
Refreshed daily. Newly registered domains are scanned and classified within the same day. Tools that go offline are flagged and pruned continuously.
Public blocklists are manually curated and typically cover a few hundred tools. Our list is machine-built from 102 million classified domains and refreshed daily with 300,000 new-domain scans, covering 16,001+ tools including the long tail that curated lists miss.
Currently 16,001+ live, deduplicated AI-tool domains, classified into 18 functional categories and 140+ subcategories. Coverage spans 210 languages globally.
Yes — and they're clearly flagged, so you stay in control. Every domain carries an ai_type flag: ai_native for products whose core value is the AI itself (ChatGPT, Midjourney, Cursor), and ai_enabled for established platforms that embedded AI features into an existing product. Blocklist and firewall customers typically filter to AI-native domains — safe to block without collateral damage — while governance and shadow-AI teams include AI-enabled platforms for the complete picture of where company data can reach AI models. One database, two lenses.
CSV, Parquet, and real-time API. Refresh cadence includes daily deltas or weekly consolidated snapshots.
Yes. Available as a standalone feed or bundled inside the full 102M-domain categorization database. Same data quality either way. Monthly and annual subscriptions available.
By classifying the web itself. We scan 300,000 new domains daily and classify each one through the same pipeline that built the 102-million-domain corpus. We also monitor app directories, product-launch platforms, developer communities, and academic repositories. If a tool has a domain, we find it — whether or not anyone's written about it.
Download a free sample CSV — real classified domains, category assignments, and metadata — and see how it fits your infrastructure.