Two platform rules changed in the last day, and neither one arrived with an announcement anyone would notice. This ai news recap july 23 2026 covers a Google Ads appeal window that closed on July 21 with no blog post and no email, and a Substack feature that now shows readers an estimate of how much of a post was written by AI.
Both stories in this ai news recap july 23 2026 are small in headline terms and large in practical terms. If you advertise on Google or publish a newsletter, one of them changes something you should handle this week.
AI News Recap July 23 2026 Lead Story: Google Ads Ended Appeals on Older Policy Decisions
Google Ads no longer lets you appeal a policy decision from inside your account once that decision is more than six months old. The change took effect on July 21, 2026, and Google recorded it as a single entry in its advertising policies change log rather than as a product announcement.
The cutoff is measured from the date of the policy decision itself, not from the day you noticed it. An advertiser looking at a November 2025 disapproval on July 20 still had an appeal button. The next day it no longer applied. For anything older than six months, the only route left is Google Ads support.
There was no blog post, no explainer video, and no thread from a Google product liaison. Trade coverage picked it up on July 22, which is why it leads this ai news recap july 23 2026.
Why the Google Ads Appeal Limit Leads This AI News Recap July 23 2026
The practical risk is the account you stopped fighting for. Plenty of small businesses have an old disapproval, a limited ad, or a suspended account they set aside because the appeal process felt hopeless. That door is now closed from inside the interface.
- Audit your account this week for disapproved ads, limited ads, and policy flags, and check the date on each.
- Appeal anything still inside the six month window now, before it ages out.
- For older decisions, contact Google Ads support and expect a slower path.
- Treat every future disapproval as time sensitive instead of something to revisit later.
This lines up with a pattern we tracked when Google Ads costs climbed earlier this year. Platform economics and platform rules are moving in the same direction, which is less slack for the advertiser, and that makes it the most expensive item in this ai news recap july 23 2026.
Substack Now Shows Readers How Much of a Post Was Written by AI
The second story in this ai news recap july 23 2026 is a labeling change. Substack launched an AI detection feature built on Pangram that estimates how much of a post, comment, or reply came from a human versus from AI assistance. It runs on text longer than 100 words and applies to content published from July 21, 2026 forward.
Chief executive Chris Best announced it in a post titled Against Claudefishing, his term for using AI to fake a human connection with readers. The scan is live on web and iOS, with Android support coming later. Writers can run the detector on their own drafts before publishing, and they can report a scan they believe is wrong.
Substack also added an optional statement where a writer can describe how they make their work. The company was candid about the limits and noted that AI detection is not settled science. Independent testing of these tools keeps landing in the same place, which is accurate but not perfect. That gap matters when a label sits next to your byline.
That caveat is the part of this ai news recap july 23 2026 worth repeating. A score is an estimate, not a finding of fact.
What Substack AI Detection Means in This AI News Recap July 23 2026
If your business publishes on Substack, the safe move is disclosure rather than avoidance. A stated process protects you from a false positive far better than silence does, because a reader who already knows how you work has no reason to treat a scan result as a gotcha.
- Add a short, honest note explaining how you use AI in your writing process.
- Scan your own drafts before publishing so the result never surprises you after it goes live.
- Keep your real voice, examples, and numbers in the draft, since no model can supply those for you.
- Do not treat a detection score as a verdict, yours or anyone else’s.
The larger signal is that platform level AI labeling is spreading fast. Google already surfaces whether an ad was made with AI inside My Ad Center, and state AI transparency rules are arriving on their own schedule. Deciding how you disclose AI use is turning into a standing business decision, and it is the quieter half of this ai news recap july 23 2026.
The Thread Running Through This AI News Recap July 23 2026
Both stories in this ai news recap july 23 2026 point at the same underlying shift. The platforms your business runs on are tightening rules faster than they are announcing them, and the window you get to respond keeps shrinking.
Google changed an appeal right in a change log. Substack attached a label to your writing without asking readers to opt in. Neither company owes you notice, and that is the real lesson. The surfaces you rent are not the surfaces you control.
The defense is unglamorous and it works. Keep your own records of every policy action taken against your accounts, check the change logs for platforms that carry your revenue, and build an owned audience such as an email list. Yesterday’s recap of the OpenAI small business program made the same point from the opposite direction. Read this ai news recap july 23 2026 as a maintenance list, not a news cycle.
AI News Recap July 23 2026: Frequently Asked Questions
These are the questions readers ask most about the two changes in this ai news recap july 23 2026.
Can I still appeal an old Google Ads policy decision?
Not from inside your Google Ads account if the decision is more than six months old. As of July 21, 2026, the in-account appeal option only covers decisions made within the previous six months. For anything older, you have to contact Google Ads support directly and work the issue through that slower channel instead.
When did the Google Ads appeal deadline take effect?
July 21, 2026. Google documented it as an entry in its advertising policies change log, with no blog post, no video, and no product liaison thread. Trade coverage surfaced it on July 22, so a lot of advertisers learned about the limit only after it was already live on their accounts.
How does the new Substack AI detection feature actually work?
Substack partnered with Pangram to scan posts, comments, and replies longer than 100 words and return an estimate of how much was written with AI assistance. It applies to content published from July 21, 2026 onward, and it is live on web and iOS now, with Android support coming later.
Will Substack punish me for using AI to write my newsletter?
Substack framed the feature as transparency rather than punishment, and there is no stated penalty. Writers can scan their own drafts before publishing, report a result they believe is wrong, and add an optional note describing how they make their work so readers know what to expect.
What should a small business do about these two changes this week?
Two things, and both come straight out of this ai news recap july 23 2026. Audit your Google Ads account for disapprovals and policy flags, then appeal anything still inside the six month window before it ages out. If you publish a newsletter, add a short disclosure about your process and scan a draft so nothing about the label surprises you later.
The bottom line of this ai news recap july 23 2026 is that two quiet changes each carry a deadline you were never sent. One closes an appeal right you may still need, and the other attaches a public estimate to writing you already publish. Both are cheap to handle this week and expensive to discover later.
Stay Ahead of the Platform Changes in This AI News Recap July 23 2026
Platform rules change quietly, and the businesses that get hurt are the ones who find out late. Our team tracks these shifts as part of ongoing SEO and analytics work so client accounts are not blindsided by a change log entry. If a policy flag, a disapproved ad, or a sudden drop in traffic is already costing you money, tell us what happened. For a daily read on stories like these, subscribe to the Demur Design newsletter in the footer below.
This recap is researched and drafted with AI, then reviewed, fact-checked, and published by Demur Design.
Sources for This AI News Recap July 23 2026
- Google Ads Help: Fix a disapproved ad or appeal a policy decision
- PPC Land: Google Ads kills appeals for policy decisions over 6 months old
- Search Engine Roundtable: Google Ads appeals limit set to 6 months
- Substack: Against Claudefishing (Chris Best)
- TechCrunch: Substack’s new tool tells you who has been writing their newsletters with AI
- Engadget: Substack is adding an AI detection feature
- The Next Web: Substack adds AI detection to fight Claudefishing
