Earlier this year OpenAI said ChatGPT had passed 900 million weekly users, and that more than 50 million people paid for a consumer subscription. The distance between those two numbers is the whole story of the past few weeks.

Advertising arrived in ChatGPT in the United States on 9 February. Canada, Australia and New Zealand followed in March, and on 11 August it reached the United Kingdom, Mexico, Brazil, Japan and South Korea. On 24 August it began serving in 31 European markets, which OpenAI calls its largest expansion to date. Jamaica is not in the rollout, and a local user should not be seeing ads today.

That is worth knowing before it arrives rather than after, because the tool a great many businesses here rely on for drafting, summarising and thinking out loud is turning into an advertising business, and the terms are already published.

What actually appears, and where

The detail matters more than the headline, and OpenAI has documented it plainly.

Ads appear on the Free plan and on Go, the low-cost tier. They do not appear on Plus, Pro, Business, Enterprise or Edu accounts, and they do not appear for accounts identified as belonging to someone under 18.

Ads can appear on Free and Go. Plus, Pro, Business, Enterprise and Edu stay ad-free, and
Jamaica is not in the rollout. (Branded graphic by PGH Consulting)

They sit below the end of a response, labelled as sponsored and visually separated from the answer itself. You are not going to find a paid placement woven into a paragraph of advice.

On what the advertiser learns, OpenAI’s position is specific: conversations are not shared with advertisers, data is not sold to them, and advertisers receive aggregated, non-identifying information such as total views and clicks. What an advertiser never receives is your chats, chat history, memories, name, email, precise location, IP address, or sensitive information such as health or political topics.

What selects the ad is a separate question, and the answer is more expansive. Targeting uses the current chat thread, your general location and your language, and if personalisation is switched on it also draws on past chats, memory and which ads you have clicked or hidden. The restraint there is placement rather than data: ads are not eligible to appear near sensitive or regulated topics, including personal health, mental health and politics. If that distinction matters to you, personalisation is a setting you can turn off.

Two different things, commonly confused: what an advertiser receives, and what can be used
to choose which ad you see. (Branded graphic by PGH Consulting)

And on the question every business owner will ask first: “Ads do not influence ChatGPT’s answers. Ads run on separate systems from our chat model, and advertisers have no ability to shape, rank, or alter ChatGPT’s responses.”

Take that seriously, then read it carefully

I think that commitment should be taken at face value. It is a specific, checkable claim about system architecture rather than a vague reassurance, and the reputational cost of being caught breaking it would be enormous.

It is also narrower than it first sounds. It says advertisers cannot change the answer. It does not say your attention is unaffected by what sits underneath the answer, which is the entire premise of advertising and always has been. If you ask which accounting package to use and a sponsored placement for one of them appears below a balanced reply, the reply is intact and your decision has still been shaped a little.

That is not a scandal. It is how every free information product you have ever used works. Free was always a business model, and what changed this month is that you can see the shape of it.

The part that concerns a business

There is a more practical worry than ad placement, and it is one this column has raised before in other forms.

When a tool is free, staff sign up individually. They use personal accounts, on personal email addresses, with personalisation on by default, and they paste in work that belongs to the business.

That was already a data question. It now also means the business has no idea what settings anyone is operating under, and no way to change them.

The fix is not a policy telling people to stop. It is giving them an account that belongs to the business, which is the same argument for approved accounts this column made when it covered AI policy and again when it covered watermarking. Ads simply make the case cheaper to win, because now there is something visible to point at.

Go US$8, Plus US$20, Business US$25 a seat, Pro US$100 or US$200. Published US prices;
OpenAI supports localised JMD billing. (Branded graphic by PGH Consulting)

What the tiers cost, and what they actually buy

The prices below are the published US figures. One useful local detail: OpenAI supports billing in Jamaican dollars, so a subscriber here is charged in JMD at a localised price rather than the US number converted. Check what your own card is actually billed before budgeting.

Go is US$8 a month in the United States. It carries ads where ads have launched, and at launch OpenAI described it as giving ten times the free plan’s messages, file uploads and image creation. Its current documentation is less specific, describing everyday text chats as unlimited on both plans with Go giving higher limits on things like image generation, uploads and data analysis. It is available here, and for a sole trader using it lightly it is reasonable value.

Plus is US$20 a month and carries no ads. For anyone using ChatGPT for real client work, this is the sensible floor.

Pro comes in two tiers, US$100 and US$200 a month, for heavy individual use. Most small businesses do not need it.

Business is US$25 per user a month, or US$20 billed annually, and adds shared workspace administration. Premium seats at US$125 a month, which became available on 25 August, add substantially more usage and remove the five-hour usage limit, which matters only if someone is genuinely working in it all day.

There is also a free way out. On the Free plan you can switch on an Ads-Free setting, which removes advertising in exchange for lower usage limits, meaning you hit rate limits more often, and reduced feature access, with OpenAI noting that some tools may be unavailable, image generation and deep research among them. Worth knowing, and worth understanding what you are giving up before choosing it.

The limits of anything published today

Two cautions.

The first is that all of this is current as at the time of writing and none of it is a promise about next year. OpenAI has said Plus, Pro, Business and Enterprise will not include ads, and its broader commitment is that there will always be a way to avoid ads, including ad-free paid tiers and a free no-ads option with additional limits. Read precisely, that guarantees some route away from advertising will exist. It does not guarantee the paid route stays at US$20.

The second is that the ad-free tiers solve the advertising question and nothing else. Paying US$20 a month does not make the model’s answers accurate, does not protect you from a confidently wrong figure, and does not remove your obligation to check anything before it reaches a client. The verification habits this column has argued for since the first week are unchanged by any of this.

Find out what plan your team is on, move client work to a paid tier, turn off ad personalisation,
and name one approved tool.(Branded graphic by PGH Consulting)

What to try this week
1. Find out which plan each person in your business is actually using, and whether they are on a personal account or one that belongs to the company. Most owners have never asked, and the answer is usually messier than expected.
2. If anyone uses ChatGPT for client-facing work, move them to a paid tier. At US$20 a month the ad question disappears and the account becomes a business tool rather than a personal one.
3. If you or your staff stay on Free or Go, open settings and turn off ad personalisation, which stops past chats and memory being used to target advertising.
4. Before switching on the Ads-Free option, check what it removes. Tighter message limits and the possible loss of image generation may or may not matter for how you actually work.
5. Write down, in one line, which AI tool is your business standard and who pays for it. Drifting onto personal free accounts is how company information ends up in places nobody chose.

Prices, plans and market availability are as published by OpenAI at the time of writing and change frequently; confirm before subscribing. Ads had not launched in Jamaica as at the time of writing. Paying for an ad-free tier removes advertising, not the need to verify what the tool tells you.

Peta-Gaye Hardy is the founder of PGH Consulting, LLC, where she helps finance and operations teams adopt AI in practical, low-risk ways. She writes the weekly AI in Finance & Business column and is based between Jamaica and the United States. Learn more at www.pghconsultinggroup.com. Follow on Instagram and YouTube @pghconsultinggroup, and connect on LinkedIn at linkedin.com/in/peta-gaye-hardy.

Disclosures: This article is informational and does not constitute investment, tax, legal, or accounting advice. Readers should consult a qualified professional before acting. Plan prices, advertising policies and market availability are as published by OpenAI at the time of writing and are subject to change; readers should confirm the current position on the vendor’s own pages before subscribing. Prices quoted are the published United States figures; OpenAI supports billing in Jamaican dollars, so local charges may differ. AI tools can produce errors, and every figure, clause, or claim they produce should be verified against a source before being shared or acted upon. The author has no commercial relationship with OpenAI or any product mentioned and was not compensated by them. All product names, logos, and trademarks are the property of their respective owners and are referenced for editorial purposes. The examples described are illustrative and do not depict any real business.