In partnership with
Stop Paying for 10 Tools. One AI Does It All.
Most e-commerce sellers are running their store across 6 to 10 separate tools — and spending more time managing software than growing their business. StoreClaw replaces your entire stack with one autonomous AI engine that monitors competitors, optimizes listings, automates marketing, and tracks real profit across Shopify, Amazon, and beyond.
It doesn't wait for you to ask. It runs 24/7 in the background, so you wake up to a full dashboard instead of a list of things you forgot to check.
Connect your store, and StoreClaw gets to work — no prompts, no complex setup, no six-app stack.
Free to start. No credit card required.
AI & Technology Intelligence
AI NEWS INSIDER
Issue #73 · September 7, 2026
THIS WEEK: Three AI vendors changed how they measure your usage in a two-week span. None of it was announced as bad news. All of it makes your AI budget harder to predict.
6 MIN READ | Sharp, actionable intelligence
Three AI Vendors Just Quietly Changed What "Unlimited" Actually Means
OpenAI, Anthropic, and Google all changed how they measure your AI usage this summer. None of the individual announcements is bad news exactly, but together, they make it a lot harder to know what you're actually paying for.
By AI News Insider Editorial · 6 min read
If your team's AI usage has felt harder to predict lately, that's not your imagination. In roughly a two-week stretch this summer, three major AI vendors each changed the rules for how much AI you actually get for your subscription, and all three moved away from a number you could simply count on your fingers.
OpenAI: "unlimited" arrives with an asterisk
On August 6, OpenAI announced that Free and Go-tier ChatGPT users would get unlimited text chats. That sounds like an unambiguous win, and for plain conversation, it is. But file uploads, image generation, and other tools stayed capped, and paid Work and Codex tiers were untouched. "Unlimited" now means unlimited for one specific type of request, not for the product as a whole.
Anthropic: the math behind a "25% increase" that will feel like a cut
Anthropic's Claude Code has been running a temporary capacity promotion since May 13 that gave users 50% more weekly usage than the original baseline. On August 29, Anthropic announced that once the promotion ends on September 13, the new standard weekly allowance will settle at 25% above the original baseline, not the boosted level. Run the arithmetic and both framings are technically true at once: 125 is 25% more than the original 100, and also about 17% less than the 150 users have had access to since May. Anyone who built a workflow around the promotional capacity will feel a real cut, even though the official framing is a permanent increase.
Google: from counting prompts to counting compute
Gemini Notebook moved on September 2 from fixed daily prompt caps, like three video overviews a day on the free tier, to a compute-based system that factors in prompt complexity, chat length, and which features you use. A short factual question now costs far less of your allowance than generating a full slide deck or video overview, and your quota refreshes every five hours instead of resetting once a day. It follows the same shift Google made to the main Gemini app back in May.
What the pattern actually means
None of these three changes is dishonest, and all of them have a legitimate technical logic: usage that scales with real compute cost is fairer to run at scale than a flat daily number. But for anyone budgeting a team's AI usage, "how much can we actually do this month" just became a much harder question to answer with a single number. The vendors are optimizing for their own infrastructure economics. Reading the fine print is now your job, not theirs.
The Bottom Line for AI News Insider Readers
Every major AI vendor is quietly moving away from limits you can count on your fingers. That's not a reason to panic, and it usually isn't bad news for heavy users specifically. But it is a reason to stop assuming last month's usage pattern will hold this month, and to read every "increase" announcement for what baseline it's actually measured against. The R.E.A.D. framework below is how to stay ahead of the next one.
You don't need a finance team to stay ahead of vendor limit changes. You need four habits, and about fifteen minutes a month.
R. Read the baseline, not the headline
Any percentage change is only meaningful next to the number it's measured against. Ask "25% more than what, exactly" before you celebrate or panic about any vendor's usage announcement.
E. Expect promotional capacity to be temporary
If a vendor's current generosity started with an announced promotion, plan for it to end. Check the actual expiration date rather than assuming it will simply continue indefinitely.
A. Audit your actual usage pattern before you need to
Most platforms show usage history somewhere in account settings. Check what your team's real monthly pattern looks like now, before a limit change catches you mid-project.
D. Diversify where it's cheap to do so
If your team relies on one AI vendor for anything mission-critical, keeping a second option warm costs little and means one company's limit change doesn't become your bottleneck.
Your Monday Morning Action
Pull up your team's AI usage history for the past month, and check whether any of your vendors have an upcoming limit change or promotion end date on the horizon. None of the three changes above were secret. All three were easy to miss.
Research
Claude Autonomously Formalizes Fermat's Last Theorem in Lean, Over 11 Days
Anthropic says the model worked through the formal proof largely without step-by-step human guidance, a notable marker for how far autonomous, sustained AI reasoning has come on genuinely hard mathematics.
Legal
Sony and Warner Sue Anthropic Over Song Lyrics, Seeking Up to $150,000 Per Song
Sony Music Publishing and Warner Chappell filed a 48-page complaint alleging Claude reproduced copyrighted lyrics; Anthropic has not yet responded in detail, and the case adds to a growing set of copyright suits testing how AI companies handle training data.
Big Tech
Google Launches Gemini 3.8 Flash and a New Enterprise Cybersecurity Model
The releases mark Google's third Flash-tier model in six weeks, alongside a dedicated cybersecurity model aimed at trusted government and enterprise customers, part of a broader push after Alphabet's longest monthly losing streak in over a decade.
Consumer
Google Begins Retiring Google Assistant on Android in Favor of Gemini
The rollout, which started September 4, is gradual and one-way, with Gemini taking over "Hey Google" triggers on phones, tablets, and Wear OS. Home speakers and cars with Google built-in keep Assistant for now.
3
AI Vendors That Restructured Usage Limits in Two Weeks
OpenAI (August 6), Anthropic (August 29), and Google (September 2).
17%
Effective Drop in Claude Code's Weekly Capacity
Once the temporary promotion ends September 13, despite being framed as a 25% increase over the original baseline.
5 hrs
New Refresh Window for Gemini Notebook's Usage Quota
Replacing a once-daily reset, effective September 2, 2026.
78%
IT Leaders Reporting Unexpected AI Pricing Charges
Tied to usage-based or consumption pricing models, per Zylo's 2026 SaaS Management Index.
Vantage (Multi-Vendor AI and Cloud Cost Visibility)
The easiest way to live the "A" in the READ framework, across every AI vendor at once
What it is: A developer-friendly cloud cost platform with dedicated LLM cost tracking, giving per-model spend breakdowns across OpenAI, Anthropic, Google, and other providers alongside your existing cloud bill.
Why it matters now: When usage limits are measured in compute rather than prompt counts, and vendors adjust the formula without much warning, a single dashboard showing what you're actually spending, by model and by team, is the difference between catching a change early and finding out on the invoice.
Who gets it: Teams running more than one AI vendor who want visibility without the heavier setup of a full FinOps platform built for Kubernetes-scale infrastructure.
• Per-model spend breakdowns across major LLM providers
• Clean, developer-friendly interface with low setup overhead
• Growing AI-specific coverage layered onto established multi-cloud cost tracking
"No one I've talked to has any clue how to handle this."
Paul Roetzer · Founder & CEO, SmarterX and Marketing AI Institute · on enterprise AI budgets going obsolete mid-year, 2026
Share AI News Insider
Know someone managing an AI budget or vendor contract? Forward them this issue before their next renewal.
SHARE & EARN REWARDS →AI News Insider · Your weekly edge in artificial intelligence
aiinsider247.com
AI Insights. Real Growth. Higher GMV, Better Profits
The difference between growing stores and stagnant ones isn't more effort. It's better insights. StoreClaw analyzes your Shopify and Amazon data, surfaces your biggest growth opportunities, and helps you increase GMV while protecting profit. Start free with bonus tokens. No credit card required.
/