AI Adoption Trends in Q1 2026: We Analyzed 67M DNS Queries/Sec and 81M HTTP Requests to Map How AI Hype Becomes Real Adoption

We tracked AI adoption across four internet layers during Q1 2026 using Cloudflare Radar data, covering 67 million DNS queries per second and 81 million HTTP requests. ChatGPT hit global domain rank #11, AI bot traffic surged 34%, and Claude overtook Google Gemini for the #2 generative AI position.

Key findings from our Q1 2026 analysis:

• The "other" TLD category (which includes .ai domains) grew 20% within Q1, from 14.7% to 17.6% of DNS queries, while .com declined from 60.8% to 58.9%
• chatgpt.com ranked #11 globally, sitting between fbcdn.net and amazon.com — the first AI-native platform in the top 15
• AI bot traffic increased ~34% over the quarter, with training-purpose crawls growing from 42% to 52% of all bot activity
• Meta-ExternalAgent emerged as the fastest-growing AI crawler (+43% within Q1), while GPTBot traffic dropped 30%
• DeepSeek shot from rank #9 to #4 among GenAI services in just three weeks during late January
• Applebot saw a 5x traffic surge by late March

Every major AI adoption report published this quarter relied on surveys. McKinsey's State of AI surveyed executives. Microsoft's AI Economy Institute polled populations across countries. Deloitte's State of AI in the Enterprise asked CIOs about their roadmaps. Their numbers are useful. But they measure what people say they're doing with AI, not what's actually happening on the wire.

Our approach was different. We pulled passive internet telemetry from Cloudflare's DNS infrastructure, covering 92 days of actual internet behavior. No surveys. No self-reporting. Just raw traffic.

The AI hype cycle has 4 measurable internet layers

Technology hype doesn't just live in headlines and analyst decks. It propagates through the internet's infrastructure in a predictable sequence, and each stage leaves a fingerprint we can measure.

We identified four layers.

The first is the domain gold rush. When a technology trend heats up, entrepreneurs and speculators register domains. DNS query patterns shift. New TLDs gain traction. This tends to run 6-12 months ahead of actual product adoption.

Second comes traffic. As products launch and users show up, domain rankings change. A platform jumping from rank #500 to the top 15 means millions of daily users. That separates working products from press releases.

Third, the bot ecosystem. AI companies need training data and they deploy crawlers to get it. The volume, targeting patterns, and purpose mix of these bots reveal which capabilities are being built right now. When training-purpose crawls spike, new model releases follow within months.

Fourth is enterprise adoption. When services reach stable, high-frequency usage patterns (weekday/weekend cycles, consistent ranking positions), hype has converted to revenue. Deloitte's latest survey found that companies with 40% or more of their AI projects in production are expected to double in the next six months. Our traffic data shows which services those production workloads are running on.

Layer 1: The domain gold rush

Cloudflare's 1.1.1.1 resolver processes over 67 million DNS queries every second. That volume makes it one of the most reliable proxies for tracking where internet activity is shifting. During Q1 2026, we watched the DNS data for signals of AI's growing presence in the domain name system.

Here's what the TLD distribution looked like:

TLD	Query share (Q1 avg)	Q1 trend
.com	61.86%	Declined from 60.8% to 58.9%
.net	12.31%	Stable
.io	1.72%	Grew from 1.65% to 1.76% (+7%)
.dev	0.61%	Stable
"Other" (incl. .ai)	~16% avg	Grew from 14.7% to 17.6% (+20%)

The .com TLD still dominates, but its share dropped from 60.8% of queries in early January to 58.9% by late March. At 67 million queries per second, a 2-point shift represents billions of daily queries moving to alternative TLDs.

The real story is the "other" category. This bucket includes .ai, .app, .xyz, and dozens of newer TLDs. It surged 20% within a single quarter. Cloudflare doesn't break out .ai separately yet, but the growth pattern aligns with the AI domain registration boom we've been tracking through our own popular technology data by category..

Why NXDOMAIN rates matter more than registration counts

Surveys can tell you how many .ai domains were registered. They can't tell you how many are actually being used. DNS data can.

10.74% of all DNS queries in Q1 hit non-existent domains (NXDOMAIN responses). One in ten queries pointed to domains that don't resolve. That's a direct measure of speculative activity, typosquatting, and recently expired registrations. For AI-related TLDs, this rate ran even higher, which suggests aggressive domain parking and speculation around AI brand names.

Domain registrations get headlines. NXDOMAIN rates tell you how much of it is real.

A geopolitical footnote

One anomaly worth flagging: .ru domains spiked to 2.16% of queries in mid-March before settling back to baseline. We've documented similar patterns in our cloud provider traffic share analysis. These spikes typically correlate with geopolitical events driving infrastructure changes or traffic rerouting.

For sales teams prospecting into AI-adjacent markets, the DNS layer is your earliest warning system. Domain speculation in a technology category tends to precede the buying cycle by 6-12 months.

Layer 2: ChatGPT enters the top 15 global domains

Domain rankings based on actual HTTP traffic paint the clearest picture of real adoption. And the Q1 2026 rankings delivered one result that would've seemed absurd two years ago.

chatgpt.com. Global rank #11.

Rank	Domain	Category
1	google.com	Search
3	cloudflare.com	Infrastructure
5	apple.com	Hardware/services
6	microsoft.com	Enterprise software
7	facebook.com	Social media
10	fbcdn.net	CDN (Meta)
11	chatgpt.com	Generative AI
12	amazon.com	E-commerce
22	bing.com	Search (AI-powered)
51	sentry.io	Developer tooling

ChatGPT now generates more traffic than Amazon. It sits between Meta's CDN and the world's largest e-commerce platform. No other AI-native platform appears anywhere in the global top 100.

Microsoft's AI Economy Institute found that generative AI adoption reached 16.3% of the world's population, up from 15.1% in the first half of 2025. Our traffic data puts a concrete shape on that number: ChatGPT's position at #11 globally requires hundreds of millions of monthly active users, and it's not sharing the spotlight. The next AI-native domain (claude.ai, perplexity.ai, deepseek.com) doesn't crack the top 100.

What's missing from the trending lists tells a story too

No AI platforms appear in the "trending rise" domain lists. That's not because growth has stalled. It's because ChatGPT has moved past "trending" entirely. When a domain stops rising fast and holds steady in the top 15, it's no longer a trend. It's infrastructure.

Bing at #22 is worth watching separately. As the backend for Microsoft Copilot, its steady ranking reflects AI-powered search growth rather than traditional Bing usage.

One more signal. Developer tooling platform sentry.io climbed to #51 globally. That's a second-order effect of AI adoption: AI-assisted coding drives more deployments, more error tracking, more developer tool usage. Harvard Business Review's February 2026 analysis found 88% of companies report regular AI use. Our traffic data shows what that 88% is actually using: overwhelmingly ChatGPT, with everything else fighting for a distant second.

Layer 3: A three-tier AI crawler economy is forming

Every major AI company needs web data. They need it to train models, power search features, and serve real-time user queries. We tracked nine major AI bot user agents across Q1 2026, and the traffic patterns reveal an ecosystem that's maturing fast.

Total AI bot traffic increased roughly 34% over the quarter, measured by a normalized index rising from 0.71 to 0.95.

Bot	Traffic share	Q1 trend	Primary purpose
Googlebot	35.0%	Stable	Mixed (search + training)
Meta-ExternalAgent	14.2%	+43%, from 11% to 15.7%	Training
GPTBot	12.3%	-30%, from 12.4% to 8.6%	Training
ClaudeBot	11.3%	Stable at 11-13%	Training
Bingbot	9.1%	Stable	Search + Copilot
Amazonbot	4.8%	Stable	Alexa AI / training
Applebot	3.9% avg, peaked 10.2%	5x surge by late March	Apple Intelligence
Bytespider	3.5%	Grew to 4.5%	Training (ByteDance)
OAI-SearchBot	2.2%	Stable	SearchGPT

Three things jumped out at us.

Meta is crawling the web harder than anyone expected

Meta-ExternalAgent was the quarter's biggest gainer. From 11% to 15.7% of AI bot traffic in 92 days. Meta has been relatively quiet about its training data strategy compared to OpenAI or Google. The crawl data tells a different story. They're consuming web content at an accelerating rate, which aligns with their Llama model development timeline. When a company's crawling ramps like this, a major model release is usually 2-4 months away.

OpenAI pulled back. The question is why.

GPTBot traffic dropped 30% across the quarter, from 12.4% to roughly 8.6%. This was the single most significant shift we observed.

Two explanations. OpenAI may have accumulated enough training data for their current model pipeline. Or the growing publisher resistance to AI crawlers, which we documented in our robots.txt blocking analysis, is having a measurable effect. Probably both. The practical result is the same: OpenAI's share of the crawl economy is shrinking while Meta's and Apple's are growing.

Apple's crawl volume says more than any WWDC keynote

Applebot surged from 2% of AI bot traffic to 10.2% by late March. A 5x increase in a quarter. Apple Intelligence launched in late 2025 with limited capabilities, and the reviews were mixed. But this crawl volume tells us they're training something far bigger than what shipped. When Apple moves quietly and at scale, you pay attention.

Training, search, and user action: the three tiers

The bot traffic breaks into three purpose tiers that reveal what the industry is building:

Crawl purpose	Share (Q1 avg)	Q1 trend
Training	45.4%	Grew from 42% to 52%
Mixed purpose	44.1%	Declined from 48% to 44%
Search	8.0%	Declined from 7.4% to 6.3%
User action	2.1%	Stable

Training crawls grew from 42% to 52% of all AI bot activity during Q1. More foundation models are coming. The industry hasn't finished building. McKinsey's State of AI survey found that 80% of high-performing companies set growth and innovation as AI objectives, not just efficiency. The crawl data shows model builders responding to that demand in real time.

Meanwhile, search-purpose crawls actually declined. AI-powered search (OAI-SearchBot, parts of Bingbot) is still a small slice of the bot economy. The real volume is in training.

What the bots are after

Industry	Bot traffic share
Retail	28.2%
Computer software	13.6%
Information technology	5.8%
Internet	5.0%
Gambling & casinos	2.8%

Retail dominates at 28.2%. Product catalogs, reviews, pricing pages, comparison content: it's structured, factual data that AI models need. If you're in e-commerce, AI bots are your second-largest traffic source after Google. The gambling number (2.8%) is surprising until you realize that casino and sports betting sites are among the most structured, data-rich websites on the internet.

Vention's 2026 AI report found that 93% of companies are already using AI in some capacity. Our bot data adds color to that number: the AI tools those companies use are themselves consuming web content at an accelerating pace, creating a feedback loop. More adoption means more crawling means more capable models means more adoption.

Layer 4: The generative AI rankings tell us who's winning

Cloudflare Radar tracks generative AI service usage daily. These rankings reflect actual user traffic, not app store downloads or press mentions. The Q1 2026 picture is a market consolidating fast at the top and churning hard below.

Rank	Service	Q1 movement
1	ChatGPT / OpenAI	#1 for 92/92 days
2	Claude / Anthropic	Rose from #4 to #2 by mid-Jan, held 85+ days
3	Perplexity	Stable #3
4	DeepSeek	Rose from #9-10 to #4 in 3 weeks (late Jan)
5	Google Gemini	Fell from #2 to #5 by mid-Feb
6	Character.AI	Stable
7	Grok / xAI	Stable
8	Suno AI	Stable
9	GitHub Copilot	Volatile, #5-9 weekday/weekend pattern
10	Doubao	Intermittent new entrant

ChatGPT's lock is tighter than you think

Number one every single day. 92 out of 92 days. No wobbles. When a product holds rank #1 in its category and ranks #11 among all global domains, displacement becomes a structural problem for competitors, not just a product quality question. You'd need to change user habits at a scale that hasn't happened in consumer tech since the smartphone transition.

Claude's rise isn't noise. It's an enterprise signal.

Anthropic's Claude moved from #4 to #2 by mid-January and held that position for 85+ consecutive days. Not a spike. A sustained shift.

Claude overtook Google Gemini, which held #2 for most of late 2025. The reason matters: Claude's focus on safety, long-context windows, and business-facing use cases is winning the segment where revenue concentration is highest. Deloitte found that only 34% of enterprises are truly reimagining their businesses around AI. The ones that are tend to pick tools built for enterprise workflows, not consumer chatbots.

DeepSeek's three-week rocket

From rank #9-10 to #4 in three weeks. That happened in late January, driven by DeepSeek's open-source model releases. The open-source angle matters here. According to NVIDIA's State of AI report, North America leads with 70% actively using AI, 27% assessing, and just 3% not using it at all. Open-source models give the "assessing" cohort a way to experiment without procurement cycles.

Whether DeepSeek sustains this position depends on trial-to-retention conversion. The daily data suggests they're holding, but the signal is noisier than Claude's steady line.

What happened to Gemini?

From #2 to #5 in about a month. Despite massive cloud infrastructure, deep integration across Workspace, Android, and Search, and distribution advantages no competitor can match, Gemini lost ground to Claude and DeepSeek simultaneously.

The market is choosing specialized AI tools over bundled ones. That's an unusual dynamic. Google has never lost a market where distribution was the deciding factor, but generative AI is apparently not that kind of market. Product quality, or at least perceived product quality, is winning.

The weekday/weekend pattern and what it reveals

GitHub Copilot bounces between #5 and #9, with clear weekday peaks and weekend drops. That's the fingerprint of enterprise adoption. A tool deeply embedded in work routines but ignored on Saturday.

Compare that to Character.AI and Suno AI. Their rankings are flat across weekdays and weekends. Consumer products. The weekday/weekend split is how you identify real B2B adoption from traffic data alone, without asking anyone a survey question.

Two new categories are forming

Windsurf AI (an AI coding assistant) and Manus (an AI agent platform) both entered the top 20 during Q1. They represent something the rankings haven't shown before: category fragmentation.

McKinsey found that 62% of organizations are at least experimenting with AI agents. Manus's appearance in the rankings is the traffic-level confirmation of that survey data. AI is fragmenting from "general chatbot" into specialized vertical tools, and the ranking data captures this as it happens. But Deloitte's enterprise survey adds a caution: only one in five companies has a mature governance model for autonomous AI agents. The adoption is outrunning the guardrails.

What the data actually tells sales and GTM teams

Cloudflare Radar shows the macro wave. DNS patterns, traffic rankings, bot behavior, service usage. But sales teams don't sell to waves. They sell to companies.

The question isn't "is AI growing?" It's: which specific companies adopted which AI tools, when they adopted them, and what those tools replaced. That's where technographic data platforms like TechnologyChecker come in, detecting 40,000+ technologies across 50M+ domains with real-time adoption and churn signals.

Here's how we'd translate the four layers into action.

If you're an SDR, the bot traffic data gives you targeting intelligence. Retail (28.2% of bot traffic) and computer software (13.6%) are the most crawled industries. Those verticals will see the most AI tool launches in the next 6-12 months. Companies already seeing heavy AI bot activity on their sites are thinking about AI strategy right now, whether they're blocking crawlers or building AI features. Digital Third Coast found that 52% of large organizations have dedicated AI adoption teams compared to 23% of small organizations. Enterprise prospects have buyers specifically tasked with AI procurement. That's who you're looking for.

If you lead GTM or RevOps, the four-layer model gives you timing. Domain speculation in a category (Layer 1) means start building content. Traffic ranking shifts (Layer 2) mean launch prospecting campaigns. Bot traffic targeting an industry (Layer 3) means those companies are 3-6 months from evaluating vendors. Stable service rankings (Layer 4) mean you're selling into an established market with clear incumbents. GitHub Copilot's weekday/weekend pattern is your template for identifying enterprise adoption in any vertical. Our technology lookup checker shows how real-time detection data can pinpoint companies at each stage.

If you run marketing, the DeepSeek and Claude stories prove that distribution doesn't guarantee market position. Google had every advantage and still dropped three positions. Content strategy should track these shifts. The audience searching for AI adoption data wants to understand dynamics, not just confirm AI is popular. Harvard Business Review found that 88% of companies now use AI regularly. Your marketing should assume AI familiarity and focus on differentiation.

The experimentation-to-production gap is real

One pattern we noticed across all four layers: there's a wide gap between AI experimentation and AI in production. Deloitte's latest data makes this concrete. Worker access to AI rose 50% in 2025, and 66% of organizations achieved productivity improvements. But only 20% are seeing actual revenue growth from AI, despite 74% aspiring to it. Only 42% believe their strategy is highly prepared.

Our traffic data maps to this gap. ChatGPT and Claude have stable, enterprise-grade traffic patterns. But below the top 5, the rankings churn weekly. Tools appear, spike, and fade. The long tail of AI services is still in experimentation mode. Only 27% of organizations review 100% of AI outputs before acting on them, per Digital Third Coast. The rush to adopt is outpacing the infrastructure to govern.

Q1 2026 AI adoption benchmarks at a glance

Benchmark	Value	Source
ChatGPT global domain rank	#11	Cloudflare Radar (our analysis)
AI bot traffic growth (Q1)	+34%	Cloudflare Radar (our analysis)
Training-purpose crawl share	45.4% → 52%	Cloudflare Radar (our analysis)
.ai-adjacent TLD growth	+20% in Q1	Cloudflare Radar (our analysis)
NXDOMAIN rate	10.74%	Cloudflare Radar (our analysis)
Most crawled industry	Retail (28.2%)	Cloudflare Radar (our analysis)
GenAI #1 consistency	92/92 days	Cloudflare Radar (our analysis)
Claude #2 streak	85+ days	Cloudflare Radar (our analysis)
Enterprise AI adoption rate	72%	McKinsey Global Survey
Global GenAI population reach	16.3%	Microsoft AI Economy Institute
Worker AI access increase	+50% in 2025	Deloitte State of AI
Orgs achieving AI revenue growth	20% (74% aspire)	Deloitte State of AI
Companies with AI agent governance	1 in 5	Deloitte State of AI
Companies regularly using AI	88%	Harvard Business Review

Our methodology

We collected data from four Cloudflare Radar API endpoints covering January 5 through April 6, 2026 (92 days, full Q1 plus the first week of April).

Cloudflare Radar draws on traffic flowing through Cloudflare's global network. That includes DNS resolution data from the 1.1.1.1 public resolver (67M+ queries/sec), HTTP request volumes, AI bot classification by user agent and declared purpose, and generative AI service rankings by traffic volume.

We used four endpoint groups:

• DNS query distribution by TLD (summary and weekly time-series)
• Global domain rankings (top 100 popular and top 50 trending)
• AI/ML bot traffic by user agent, crawl purpose, and target industry
• Internet services rankings filtered to "Generative AI"

All trend percentages compare the first week of Q1 to the last available data point. Bot traffic was analyzed by both user agent identity and declared crawl purpose. GenAI service rankings used daily position data to calculate streak lengths and transition points. We cross-referenced signals across layers to identify patterns that wouldn't be visible in any single data set.

What Cloudflare doesn't see: traffic that doesn't flow through its network, DNS queries to other resolvers (Google Public DNS, ISP resolvers), and .ai TLD breakouts (we track it within the "other" bucket). Bot user agents can be spoofed, though major AI companies have strong incentives to identify their crawlers honestly for robots.txt compliance. Our SSL certificate transparency research follows a similar methodology, using infrastructure-level data to draw adoption conclusions.

Frequently asked questions

What are the AI adoption trends in 2026?

Based on our Cloudflare Radar analysis covering Q1 2026, AI adoption has moved past experimentation for the top platforms. ChatGPT ranks #11 globally among all domains. AI bot traffic grew 34% in a single quarter. The GenAI service market has consolidated into a clear top 5: ChatGPT, Claude, Perplexity, DeepSeek, and Google Gemini. Training-purpose crawls grew from 42% to 52%, which means model development is accelerating, not slowing. McKinsey's survey data (72% enterprise adoption) and Deloitte's findings (worker AI access up 50%) align with what we're seeing in the traffic.

What are AI adoption trends by industry?

Our AI bot traffic analysis shows retail leads all industries at 28.2% of crawler traffic, followed by computer software (13.6%), information technology (5.8%), and internet services (5.0%). Gambling and casinos round out the top 5 at 2.8%, likely because their sites are unusually structured and data-rich. These numbers reflect where AI companies are directing their training resources. The industries getting crawled most heavily will see the most AI-powered products entering the market in the next 6-12 months.

How does AI adoption vary by country?

Our Cloudflare Radar analysis focuses on global aggregates, but two geographic signals stood out in Q1. The .ru TLD spiked to 2.16% of DNS queries in mid-March, reflecting geopolitical infrastructure shifts. And Doubao's intermittent appearance in the GenAI top 10 signals Chinese AI platforms beginning to expand globally. NVIDIA's State of AI report found North America leads with 70% actively using AI, 27% assessing, and just 3% not using it. Most top-ranked AI platforms (ChatGPT, Claude, Perplexity) are headquartered in the US with predominantly North American traffic patterns, but DeepSeek and Doubao suggest that's changing.

What are the latest AI growth statistics?

From our Q1 2026 data: AI bot traffic grew 34% in 92 days. Training-purpose crawls reached 52% of all AI bot activity. ChatGPT holds global rank #11 with more traffic than Amazon. Meta-ExternalAgent grew 43% within the quarter. Applebot surged 5x. On the service side, Claude took #2, DeepSeek reached #4, and Gemini dropped to #5. MIT Sloan Management Review's Thomas Davenport and Randy Bean identified potential AI bubble deflation as a 2026 theme. Our traffic data shows the opposite at the infrastructure level: usage is still accelerating. The disconnect might be that public market sentiment is cooling while actual adoption keeps climbing.

Where can I find an AI adoption graph for 2026?

This article contains six data visualizations covering Q1 2026: DNS query share by TLD, global domain rankings, AI bot traffic by user agent, crawler purpose distribution, generative AI service rankings, and a GTM strategy summary. Each uses Cloudflare Radar data from January 5 to April 6, 2026. For additional technology tracking data, see our industry statistics insights and technology category breakdowns.

What do recent reports say about AI adoption trends in 2026?

The major reports this cycle include McKinsey (72% enterprise adoption, 62% experimenting with AI agents), Deloitte (worker AI access up 50%, but only 20% achieving revenue growth), Microsoft (16.3% global population using GenAI), NVIDIA (70% of North American companies actively using AI), and HBR (88% regular use). Our research adds the infrastructure dimension. Real traffic flows, actual crawler behavior, daily service rankings. Where surveys tell you what companies plan to do, traffic data shows what they're actually doing today.

What percentage of companies are using AI?

It depends on how you define "using." Vention's 2026 report puts it at 93% of companies using AI in some form, with 80% using it directly in operations. McKinsey's Global Survey says 72% have adopted at least one AI capability. HBR says 88% report regular use. Our Cloudflare Radar data adds a different dimension: ChatGPT alone ranks #11 globally for web traffic, ahead of Amazon. But Deloitte's data reveals the nuance beneath those headline numbers. Only 20% of organizations are actually generating revenue from AI, despite 74% aspiring to. The gap between "using AI" and "getting business value from AI" remains wide.

How fast is the AI market growing?

Our infrastructure-level data shows AI-related internet activity growing at roughly 34% per quarter (Q1 2026), measured by AI bot traffic volume. The "other" TLD bucket that includes .ai domains grew 20% within Q1 alone. Training-purpose crawls jumped from 42% to 52% of all AI bot activity, signaling that model development is still accelerating. Microsoft's AI Economy Institute found that global generative AI adoption reached 16.3% of the world's population, up from 15.1% in H1 2025. On the services side, our data shows the top 5 GenAI platforms consolidating their positions while the long tail churns rapidly.

Which AI has the highest market share?

By actual traffic volume, ChatGPT is the clear leader. It held the #1 position among generative AI services for all 92 days of Q1 2026, and ranked #11 among all global domains. No other AI service comes close. Claude (Anthropic) held a stable #2 for 85+ consecutive days. Perplexity locked in #3. DeepSeek rose to #4. Google Gemini fell to #5. The important detail: no other AI-native domain besides chatgpt.com appears in the global top 100. The gap between #1 and #2 in generative AI is far wider than the ranking numbers suggest.

Is ChatGPT losing market share?

Not according to our traffic data. ChatGPT held the #1 position among generative AI services every single day of Q1 2026, and its global domain ranking at #11 puts it ahead of amazon.com. What's shifting is the competition below it. Google Gemini fell from #2 to #5 during Q1, losing ground to Claude and DeepSeek. But ChatGPT's position hasn't weakened. If anything, its move into the top 15 global domains suggests continued growth at a scale the other AI services haven't reached.

Why do 85% of AI projects fail?

The often-cited failure rates (85%, 90%) come from Gartner estimates about enterprise AI projects not reaching production. Our traffic data offers a partial explanation. Below the top 5 GenAI services, the rankings churn weekly: tools spike, plateau, and fade. This pattern mirrors the experimentation-to-production gap that Deloitte quantified: 66% of organizations achieved productivity improvements from AI, but only 20% are generating revenue. Only one in five companies has a mature governance model for AI agents. The problem isn't that AI doesn't work. It's that most organizations adopt faster than they can operationalize. They're running before they've figured out how to walk.

Is AI more hype than reality?

Both, depending on which layer you're measuring. At the DNS level (Layer 1), there's genuine speculation: 10.74% of DNS queries hit non-existent domains, and the "other" TLD bucket (including .ai) grew 20% in a quarter. That's hype activity. But at the traffic level (Layer 2), ChatGPT at global rank #11 is not hype. That requires hundreds of millions of active users. At the bot level (Layer 3), training crawls growing from 42% to 52% shows real infrastructure investment, not just announcements. And at the service ranking level (Layer 4), the top 5 GenAI platforms held stable positions for months. MIT Sloan's Thomas Davenport flagged "AI bubble deflation" as a 2026 theme. Our data says the bubble may be deflating in public markets, but the underlying usage is still accelerating.