6 min read
What happens to your business when the companies that make your AI tools also get to write the rules those tools follow? That is not a hypothetical anymore. This month, Anthropic, Google, and OpenAI confirmed they are building a shared AI standards body, aiming to launch by the end of 2026 or early 2027. If you run a one-person business on top of ChatGPT, Claude, or Gemini, this is your industry deciding its own guardrails, and it is worth understanding before you are asked to trust the result.
Let me be upfront: I think this is mostly good news for solo owners, with a real catch you should watch. The temptation is to file “AI safety standards” under things that only matter to big tech lawyers. That would be a mistake. The tools you now depend on to write proposals, answer customers, and run your admin are about to operate under a rulebook their own makers are drafting. That affects reliability, price, and choice, which are three things a business of one cannot ignore.
What is actually being built, minus the press-release gloss
Strip away the language and here is the signal. The three companies plan to stand up an independent-sounding organization that would support third-party testing of AI models before they ship, set protocols for reporting safety and security incidents, define voluntary safety commitments, and lay out who counts as a qualified auditor. It may even run its own model testing. This is not starting from zero; it builds on the Frontier Model Forum these firms created back in 2023, and you can read that group’s own framing on the Frontier Model Forum site.
The word doing the heavy lifting is “voluntary.” There is no government forcing this. An earlier attempt at a public-private partnership stalled, so the companies are choosing self-regulation instead. That is the whole story in one sentence: the industry could not get an outside referee, so it is hiring its own.
Why the timing should get your attention now
You might think a rulebook that arrives in 2027 is a 2027 problem. It is not, because the behavior it is meant to govern is already in your business. AI tools now take actions on your behalf, not just answer questions. If you have ever let an AI browser act on your accounts, you have already handed real authority to a system whose safety testing you cannot see. Recent incidents where AI models reached across the internet and affected other organizations, some of them not fully disclosed, are exactly the kind of event this body is meant to catch. The standards are being written because the risk is present tense.
There is a second reason to care about timing. When an industry sets its own standards, the cost of meeting those standards eventually shows up in pricing and in which tools survive. Decisions being made now will shape what you pay and what you can choose from a year from now.
The catch worth naming out loud
Here is where healthy skepticism earns its place. Critics warn that a standards body run by the three largest AI companies could quietly box out smaller and open-source competitors, by defining “safe” in ways only the giants can afford to meet. That is not a conspiracy theory; it is the normal gravity of any club that writes its own membership rules. For a solo owner, the risk is not dramatic. It is slow. It looks like fewer cheap alternatives and more lock-in to a handful of big platforms over time. The trend of AI quietly consolidating is one we have tracked before, including as it moves off the cloud and onto devices.
Three ways to position your business
You do not control any of this, but you do control how exposed you are. Three sane postures, with honest trade-offs.
Option one: lean in and standardize. Pick one major provider, learn it deeply, and accept that the biggest players are also the ones most likely to clear whatever bar this body sets. The upside is stability and support. The trade-off is dependence on a single company’s roadmap and pricing.
Option two: stay deliberately portable. Keep your prompts, your data, and your workflows in formats that are not welded to one tool, so you can switch if a provider raises prices or drops a feature you rely on. The upside is freedom. The trade-off is that you do a little more maintenance and never go quite as deep on any one platform. Our comparison of ChatGPT vs Claude vs Gemini is a good starting point for keeping real options open.
Option three: watch and wait. Change nothing yet, but pay attention to whether your tools start advertising compliance with the new standards, and whether that correlates with price increases. The upside is you avoid reacting to a rulebook that does not exist yet. The trade-off is that “wait” can quietly become “got caught flat-footed” if you stop watching.
What to ignore
Ignore the fear that safety standards will make your everyday tools worse or lock you out. For a solo owner writing captions and drafting emails, day-to-day use will feel the same. Ignore, too, the breathless framing that any single announcement changes everything overnight. It does not. What changes is the direction of travel, and direction is a slow thing you steer against, not a fire you put out. The genuinely urgent AI risks for a small business remain the boring ones, like the fact that AI has made scams cheap to personalize. Guard against those first.
Your next 90 days
Do one concrete thing this quarter: run a five-minute audit of your AI dependence. Write down every tool your business would genuinely struggle without, note who owns each one, and mark which of your critical data lives inside a single provider you could not easily leave. That list is your real exposure. You do not have to act on it today. You just have to be able to see it, so that when the standards land and the pricing shifts, you are reading a map you already drew instead of scrambling.
The companies building this rulebook are not villains, and self-regulation is not automatically a stitch-up. But “trust us, we wrote our own rules” is a claim a careful business owner examines rather than swallows. Watch what these standards actually require, watch whether smaller tools can still compete under them, and keep enough portability that your business is never the one with no way out. That posture costs you almost nothing today, and it is the difference between being a customer and being captured.
