Cloudflare charges crawlers, Google widens AI: July 2 brief

The free ride AI companies have been taking on your website content may be ending, and Cloudflare made the biggest move toward that today. AI traffic is shifting from unpaid scraping (bots reading your site's content in bulk, for free) toward priced access. That hits publishers, niche site owners, software businesses with valuable data, and any owner counting on search traffic. Google also widened its AI product surface, while AI coding assistants kept turning into a budget line. Fast week.

Cloudflare turns AI access into a billable request

Cloudflare, the company that sits in front of a large share of the world's websites and manages their traffic, announced its Monetization Gateway on July 1. It opened a waitlist for customers who want to charge for web pages, datasets, or the connection points other software uses to pull their data (APIs). The payments settle through x402, an open standard that lets software pay for content automatically, per request. Cloudflare says it will verify payment before traffic ever reaches your actual site.

This matters because the old content bargain is cracking. Search crawlers, the automated programs that read websites, used to send traffic back in exchange. AI agents can read once, summarize, and move on. Cloudflare says AI crawlers already request content anywhere from hundreds to tens of thousands of times for every visitor they send back. That is a brutal exchange for anyone funding content with ads, leads, or subscriptions.

The move is not to block every bot on sight. Too blunt. First separate three buckets: pages that need to be found, assets worth protecting, and data other software might pay to use. A buying guide may need to stay open so search engines and AI can find you. A proprietary dataset, calculator, or premium data feed probably does not.

This is the same pressure behind our AI Overviews briefing and the AI recommendations piece. Getting discovered is getting harder. Owning the underlying data is getting more valuable.

Google packs more work into Gemini and NotebookLM

Google used its June AI roundup, published July 1, to group a long list of product updates: Gemini 3.5 Live Translate, Gemini 3.5 Flash with the ability to operate a computer on your behalf, Gemma 4 12B for running AI locally on ordinary 16GB laptops, NotebookLM upgrades, Google Finance AI features, and Gemini Spark updates.

The piece most useful to a business owner is NotebookLM, Google's tool for asking questions across a pile of your own documents. Google says it added stronger reasoning, a secure environment for running code, and the ability to generate charts, spreadsheets, and slide decks. Access is global for Google AI Ultra subscribers and specific Workspace accounts. That makes NotebookLM less like a note tool and more like a research assistant for messy inputs.

The move is to test it on internal research before anything customer-facing. Feed it sales calls, support complaints, survey notes, and competitor pages. Ask for the spreadsheet and chart first, then use a human to decide what is true and useful. Do not push generated slides straight to customers. Rough edges still show.

The broader read: Google is making AI a default layer across Search, Workspace, Finance, Android, and home devices. If your plan for being found still assumes a customer clicks ten links and compares manually, it is behind. Our AI search topic page tracks the same shift from search result to answered task.

GitHub adds a lower-cost open model to Copilot

GitHub made Kimi K2.7 Code generally available in Copilot on July 1. Copilot is GitHub's AI coding assistant, and this is the first open model (one anyone can download and inspect) offered as a selectable option inside it, hosted by GitHub on Microsoft Azure and billed by usage at the provider's list price.

Rollout starts with Copilot Pro, Pro+, and Max. GitHub says company administrators must switch the option on before their teams can use it. That admin toggle matters. The cheap model still has to clear your security and data rules.

The read for an owner is simple: AI is no longer one model choice. It is a routing problem, deciding which jobs go to the cheap tool and which justify the expensive one. Cheap models handle small edits and routine work. Expensive models get the high-risk changes. The same pattern is coming to marketing: one assistant for drafts, another for analysis, another for automations.

Cursor's CursorBench 3.1 page shows why price now belongs in every AI test. It compares quality scores against average cost per task across models such as GPT-5.5, Sonnet 5, Gemini 3.5 Flash, Kimi K2.7 Code, and GLM 5.2. The standout detail is not one winner. It is the spread: higher scores can cost many times more per task.

For small teams, the move is to stop buying the most expensive model by habit. Make a two-column test: is the quality good enough, and what does each completed task cost. That discipline applies just as much to AI writing tools and content work as it does to code.

Worth watching

Meta is reportedly exploring a cloud business selling raw computing power and model access. If that happens, model access gets more fragmented, but compute-heavy tools may get cheaper.

Venice AI raised a $65 million Series A funding round at a $1 billion valuation, according to TechCrunch. Privacy-first AI keeps pulling demand, but treat unfiltered tools carefully around your brand and customer data.

HubSpot's reported Warmly deal is worth tracking, but we could not verify the primary source cleanly today. If confirmed, it points to customer-tracking software moving closer to identifying your website visitors and timing sales outreach with AI.