There is no single blockbuster ChatGPT announcement from the past 30 days that reshapes how your firm gets found. The more useful story is a through-line: over the last year, OpenAI has steadily changed how ChatGPT surfaces, cites, and shops for brands, and third-party trackers show that the citations feeding those answers move in ways no brand controls. For professional-services firms in Australia and the UK, that shift matters more than any one release.
What changed
Start with the most directly visibility-relevant feature, and note that it is now background rather than breaking news. On 24 November 2025, OpenAI introduced a conversational buyer's-guide flow, rolled out to all logged-in users, as reported in Search Engine Journal's coverage of ChatGPT shopping and AI citations. That is roughly eight months old now, so treat it as the established backdrop: ChatGPT increasingly answers "which provider should I use" questions inside the chat, pulling from web sources rather than sending people to a results page.
More recently, the changes have been about the models themselves. Around 26 June 2026, OpenAI split GPT-5.6 into three tiers (Sol as the flagship, Terra as the balanced option, and Luna as the lightweight one) and retired GPT-4.5 from ChatGPT at about the same time, per the OpenAI release notes. These are capability releases, not visibility features. Any effect they have on how your firm gets surfaced is indirect: a different model can reason over and phrase citations differently, but OpenAI has not published a direct link between the tier split and how brands are chosen or named.
The volatility is the part firms feel most. The measurement platform seoClarity reported notable drops in ChatGPT citation volume around 8 March 2026 and again around 19 April 2026, followed by a rebound in May 2026, in its analysis of ChatGPT citation decline. Two cautions matter here. First, these are third-party measurements, not statements from OpenAI. Second, seoClarity frames the swings as platform-level shifts in citation behaviour rather than a verdict on any individual brand's content. The takeaway is not a number to chase but a principle: AI citation is volatile and not directly controllable.
What it means for your firm being found
If you run an accounting, law, health, or consulting practice, or an agency serving them, the practical reading is straightforward. ChatGPT is becoming a place where prospective clients ask for a shortlist, and the sources it cites can rise or fall sharply on a timeline you do not set. A strong month of visibility can be followed by a weak one for reasons that have nothing to do with your website.
For Australian and UK firms specifically, this compounds an existing problem. Local intent (a Sydney tax specialist, a Manchester employment lawyer) already gets thinner AI coverage than large US-centric queries, so a platform-wide citation dip can hit regional professional-services firms harder. You cannot lobby the model, but you can make yourself the kind of source that clear, well-structured web content tends to favour when citations do flow. That is Amina's north star in practice: built to be found, in search and in AI.
What to do now
Do one thing: run a baseline check of whether ChatGPT can currently find, name, and correctly describe your firm for the questions your clients actually ask. Ask it directly ("who are the best employment lawyers in Brisbane for small business") and record what it cites, what it gets wrong, and whether you appear at all. That baseline is what makes the volatility legible, because you cannot tell a platform-wide dip from a your-firm problem without it. Amina's AI Visibility Check is built to run exactly this diagnostic across the pages and prompts that matter to your market.
FAQ
Does OpenAI's GPT-5.6 tier split change how my firm is cited?
Not in any way OpenAI has confirmed. The Sol, Terra, and Luna tiers released around 26 June 2026 are capability changes. Any effect on how brands are surfaced is indirect, so treat model releases as context, not a lever you can pull.
Why did my ChatGPT visibility drop earlier this year?
It may not be your site. seoClarity measured broad citation-volume drops around 8 March and 19 April 2026, with a rebound in May, and described them as platform-level shifts. These are third-party measurements, so read them as evidence of volatility rather than a score on your content.
Is ChatGPT shopping relevant to a services firm, not a retailer?
Yes, in principle. The shopping research flow launched on 24 November 2025 as a buyer's-guide experience, and the same "recommend me a provider" behaviour increasingly applies to services, though its direct impact on professional-services citations is not something OpenAI has quantified.
