While you were busy kicking off your 2026 game plan, AI platforms were making big ecommerce moves, introducing ads, giving us a glimpse of the agentic future, and more.

If we had to boil down Q1 2026 into three core themes, they would be:

1. AI search followed a familiar path: From shopping to ads, AI platforms kept extending their reach into areas long dominated by traditional search.
2. Agent-first went mainstream: Both emerging entrants and established leaders made waves on the AI agent front—not just with new tools, but with new infrastructure built for bots.
3. Consumer demand surged and shifted: Claude's breakout quarter and Clawdbot's virality showed that platform loyalty in AI is still very much up for grabs.

Here’s what happened in AI search last quarter—and what it means for your strategy.

AI shopping carts came into focus

What happened

In the wake of ChatGPT’s Instant Checkout launch last September (which OpenAI has since cooled on in favor of app-in-chat purchases), both Google and Microsoft made their own ecommerce waves.

Google unveiled a new open standard for agentic commerce—Universal Commerce Protocol (UCP)—on January 11 that allows for AI agents to make direct purchases in Google AI Mode and Gemini. UCP, developed alongside Shopify, is not to be confused with OpenAI’s own open-source protocol—Agentic Commerce Protocol (ACP)—developed alongside Stripe.

Speaking of which….

Microsoft launched Copilot Checkout on January 8, running on, you guessed it, ACP.

What it means

ChatGPT may have detoured from Instant Checkout, but the ecommerce land grab is still very much on.

AI platforms are gearing up for a near future where AI user agents aren’t just answering questions about your products, but buying them on your behalf. In the meantime, AI search users will find it easier to spend dollars without hopping tabs on a merchant’s website.

If you sell things online, get familiar with different platforms’ approaches and protocols. Talk to your ecommerce team. Understand which surfaces your customers are on and which checkout experiences they're likely to encounter (hint: Scrunch’s Shopping feature will help).

AI ads entered the chat

What happened

OpenAI announced it would start testing ads in ChatGPT on January 16 (and was promptly roasted by rival Anthropic with a series of Super Bowl ads).

The ChatGPT maker kicked things off with its first ads on February 9. Retail behemoths like Target, Williams-Sonoma, Ford, and Mazda reportedly paid a $200,000 minimum buy-in for the privilege of trying things out on a cost per thousand impressions basis.

Since then, OpenAI has snagged one of Meta’s former advertising mavens and reportedly decided to extend its pilot.

What it means

If brands start spending serious money to advertise on AI platforms, they’ll demand audience and ROI transparency.

It’s a fairly safe bet that we’ll get more visibility into prompt volumes and similar metrics directly from ChatGPT (and other platforms that follow suit). We’re seeing the start of this already with Bing adding AI performance to its webmaster tools.

In other words: The AI black box will begin to crack open.

Clawdbot took the internet (and GitHub) by storm

What happened

Technically Clawdbot (then Moltbook after Anthropic’s lawyers got involved, now OpenClaw) launched in late 2025, but it didn’t go viral until late January, going on to amass hundreds of thousands of stars on GitHub.

Long story short: Open-source AI assistant that runs on your own infrastructure (local machine, a Mac mini, or a virtual private server), instead of the cloud and connects to your messaging apps.

Plus a social network where AI assistants (or people pretending to be AI assistants) can interact and plan world domination?

What it means

Along with Claude Code, Clawdbot threw fuel on the “agentic” fire by showing how AI can truly become an autonomous agent versus just a Q&A interface.

Our take? Whether it’s humans doing the searching or their AI assistants, your brand and product will still need to be discovered easily, represented accurately, and described favorably in AI search results to compete.

In short: You need to deliver a web experience that makes it easy for AI bots to retrieve the info they need (and eventually transact).

AI bots started getting the red carpet treatment

What happened

Bot-first optimization made headlines with Google opening up WebMCP for preview on February 10 and Cloudflare releasing Markdown for Agents on February 12.

The people who run a significant portion of the internet are waking up to the idea that they need to treat AI user agents as first-class visitors with their own dedicated web experiences.

We’re in the early innings of establishing industry standards for how to optimize for AI bots—expect more news on this front throughout the year.

What it means

It’s important not to confuse a vision for how AI might work in the future with how it actually works today.

For example, Cloudflare’s release only activates for agents that explicitly request markdown. As of today, that’s primarily Claude Code and OpenCode, not mainstream AI surfaces like ChatGPT, Gemini, or standard Claude.

Both consumer AI and coding agents typically look for information via “regular” web search and consume regular HTML from pages they find in search results. That’s where Scrunch’s Agent Experience Platform (AXP) comes in, delivering agent-optimal HTML while keeping your human website the same, minus any technical heavy lifting.

Claude got a major shot in the arm (including from the US government)

What happened

This story had been building for months, but things really came to a head in late February. TLDR: After Anthropic refused to back down on restrictions for how the US Department of Defense could use its models, the DoD labeled Anthropic a “supply chain risk” and signed a deal with OpenAI.

The upshot? Anthropic became the first American company to ever publicly be called a supply chain risk and Claude immediately started shooting up the AI leaderboards.

Claude beat ChatGPT for daily active users and saw a massive 1,487% increase in sessions in early March.

What it means

Claude’s massive spike in popularity following the events with the DoD (the rapturous response to Claude Cowork and Claude Code certainly didn’t hurt either) is a helpful reminder for brands:

It pays to approach AI search with a multi-model mindset.

If you weren’t keeping an eye on your brand performance in Claude before it reached No. 1 in the Apple app store, now’s the time to start.

Google cracked down on AI slop

What happened

Google rolled out a spam update on March 24 aimed in part at scaled content abuse.

Put another way: Sites that rely on AI to generate vast quantities of low-quality content to game search results may see lower rankings or removal from results altogether.

Fun fact: This was the fastest spam update in Google history, taking less than 20 hours.

What it means

The relationship between traditional search ranking and AI search performance is not one-to-one, but tanking in SERPs certainly won’t do you any favors (especially across Google’s AI surfaces).

The takeaway for brands? Optimizing high-quality content for AI beats generating low-quality content with AI. Google isn't punishing brands for using AI to polish or improve content—they're punishing mass-produced slop.

Our advice: Make authoritative content easier for AI to consume and cite instead of cranking out surface-level SEO spam and hoping that something sticks.

Scrunch product launches in Q1 2026

AI platforms weren’t the only ones busy in Q1.

Scrunch was building to help brands map their websites, stay on top of AI shopping, scale citations, expand platform coverage, and cut down on costs, including:

Site Maps

We launched our Site Maps feature, allowing you to map your website and get a page-by-page view of AI search performance.

Page-level drill-downs, AI page auditing, automated content optimization, and AI-friendly content delivery—all in one workflow.

Shopping

We launched a major upgrade to our Shopping experience, giving you full visibility into product performance, competitive landscape, retailer share, and more in AI search.

See how your SKUs are showing up, which competitors are winning your shelf space, which retailers are capturing your AI-driven clicks, and which prompts are behind the results.

Noble and Stacker partnerships

We announced our partnerships with Noble and Stacker, making it easier to identify citation opportunities and take action at scale.

Secure brand mentions in cited sources via Noble, turn owned content into earned placements across publishers via Stacker, and track performance via Scrunch.

Grok support

We launched support for Grok, bringing the total number of AI platforms supported in Scrunch to nine (including ChatGPT, Gemini, Perplexity, Google AI Overviews, Google AI Mode, Claude, Copilot, and Meta AI).

All the real-time monitoring and optimization we offer for other AI surfaces, now for Grok.

Topic Prompt Optimizations

We launched our Topic Prompt Optimizations feature, designed to help you find the ideal number and mix of prompts for each topic you care about.

Identify which prompts deliver unique value and cut the ones that don’t to reduce cost, noise, and redundancy.

Stay tuned for our Q2 2026 digest.

Get started with Scrunch

Start reaching more customers on AI platforms. Spin up a free account or see Scrunch in action today.