ChatGPT goes unlimited, Time targets AI: August 7 brief
ChatGPT removes text limits for free users, Time tests ads for AI crawlers, and Suno adds tracking to generated music.

Some links below are partner links. If you buy through one, The Memo earns a commission at no extra cost to you. How we make money.
In this briefing
ChatGPT is about to give free users unlimited text conversations, raising the baseline for what customers and small teams can do without another subscription. Time is testing ads meant for AI assistants rather than people, while Suno is making generated music easier to trace. The common thread is access with tighter accountability.
OpenAI removes ChatGPT's free text ceiling
OpenAI said on August 6 that ChatGPT's free plan will get unlimited text chats starting next week. Free and Go users will also receive GPT-5.6 Luna as their standard model and a Think button that lets the system spend more time on a difficult question. Limits will remain on file uploads, images, and other extras, and OpenAI says abuse controls still apply.
The same OpenAI announcement brings a response-effort slider to paid Plus and Pro accounts. That setting controls how long ChatGPT works before replying. OpenAI also reports that, in its own test of fact-heavy financial, medical, and legal questions, answers containing an error were 62% less common with Luna and 68% less common with Sol than with the previous GPT-5.5 Instant model. Those are company test results, not a promise that either model is mistake-free.
For an owner, unlimited text removes the easiest reason to buy a paid account for occasional writing, planning, or customer research. Paid plans now need to earn their place through heavier features, not a larger message allowance. It also means more customers can use ChatGPT throughout a buying decision, making our guide to how AI recommends businesses more relevant outside early adopters.
Your move
Time starts selling ads to AI readers
Time is serving some AI search crawlers a different version of its pages, complete with sponsored material that ordinary visitors do not see. Software developer Vincent Schmalbach documented the setup on August 5. His tests found that ClaudeBot, PerplexityBot, and OpenAI's search crawler received a short text version, while browsers and Google's search crawler received the normal website.
The machine-facing pages included sponsored FAQs from Ally Bank and the Project Management Institute. They were labeled as sponsored inside the text supplied to the crawler. The concern is one step later: if an assistant uses that material in an answer, the person asking may not see how paid content entered the chain.
This is an early ad format, not evidence that buying one of these placements will change what ChatGPT says. Small businesses should watch the boundary between earned mentions and paid distribution before putting money into it. First, make the facts you control easy to retrieve through complete product pages, current business profiles, and clear policies. Our AI search traffic briefing explains why being quotable now matters even when the searcher never visits your site.
Suno makes AI music easier to identify
Suno said on August 6 that it is rolling out ways to identify songs made on its service and will add audio watermarking and fingerprinting in the coming weeks. A watermark is a hidden marker inside a file. A fingerprint is a digital pattern platforms can compare against known audio. Suno says both are designed to survive tampering without changing the listening experience.
The company's policy update also promises new download limits aimed at mass distribution to streaming services. Suno did not publish the exact thresholds. Its revised rules more clearly ban copied songs, unauthorized voices, spam, fake activity, and attempts to evade controls with bots.
Any team using generated music in ads should preserve the project record, source file, license terms, and approval date. Identification marks help platforms trace a track, but they do not prove that your business has permission to use every voice, lyric, or uploaded sample inside it. Apply the same consent discipline used for AI voice in advertising.
Worth watching
- Snapchat published humor guidance based on a survey of more than 6,000 people aged 13 to 28. Useful creative input, but its platform-funded findings should not replace testing with your own customers.
- OpenAI's reported smart speaker remains a hardware rumor. Do not plan customer service or commerce around it until there is an official product.
- Time's crawler ads will matter more if assistants disclose paid source material or publishers release outcome data. Neither has happened yet.
About The Memo
The daily brief on AI and marketing. What changed in AI tools, search, ads and growth, why it matters, and the move to make this week. How we work
The daily Memo
One short email each morning when something in AI or marketing actually changes, with the move to make.
Free. Every morning. Unsubscribe anytime.
Keep reading
Related briefings

Google cuts AI costs, Substack scans copy: July 22 brief
Google lowers the cost of fast AI work, Substack lets readers scan for AI-written copy, and OpenAI starts a small-business program.

OpenAI voice, Grok pricing: the July 9 brief
OpenAI made voice AI more usable, xAI cut model costs, Cloudflare made site launches easier, and Google Search demand stayed loud.

Copy.ai alternatives 2026: 3 clearer choices
Copy.ai alternatives for small marketing teams, compared on current price, brand control, search research, team access, and the work each handles well.