🎧 Prefer to listen?

I woke up last week to find that the AI model I’d been using for my client automation had been pulled. Not because it broke — because the government decided it needed to review it first. If you’re using Claude, ChatGPT, or any tool built on top of them, the same thing could happen to you.

I already wrote about why the Anthropic vs. OpenAI rivalry doesn’t matter anymore — both companies are now stuck in the same regulatory bottleneck. But that post focused on strategy. This one’s about the practical reality: which of your tools are actually at risk, what the government approval process looks like, and what you can do right now to protect your workflows.

What’s actually happening with AI model approvals

Here’s the situation as of July 2026. The US government has started requiring approval before frontier AI models can be released to the general public. Anthropic’s Mythos model has been stuck in a limited preview for months — available to some enterprise customers, but not broadly released. Now OpenAI’s GPT 5.6 is facing the same process, with the government approving its release “customer by customer.”

This isn’t a ban. The models still exist. But the approval process means that new capabilities — the ones that would normally roll out to everyone at once — are now drip-fed through a government bottleneck. If you’re a solo builder who relies on having access to the latest models, this changes when and how you get them.

The bigger issue, as GMU fellow Dean Ball laid out, is that nobody seems to know what the government is actually testing for. There’s no published checklist, no clear timeline, and no appeal process. The US government doesn’t have the technical capacity to evaluate frontier models the way independent researchers do — which means the approval process could be slow, arbitrary, and inconsistent.

Which tools are actually at risk

Not everything is affected equally. Here’s how I’d break it down for non-coders:

High risk: Tools that depend on a single frontier model. If you’re using a tool that runs exclusively on Claude or exclusively on GPT, and that tool relies on the latest model version, you’re most exposed. When Anthropic’s Mythos got pulled, every tool built specifically on Mythos lost access overnight. The same could happen with GPT 5.6.

This includes things like Claude’s agent features — if Anthropic can’t release new model versions, Cowork’s capabilities freeze at whatever was approved before the hold.

Medium risk: Multi-model platforms. Tools like Make.com, n8n, and Zapier that let you connect to multiple AI providers are safer. If one model gets held up, you can route to another. I covered this in my model-agnostic strategy post — but even multi-model platforms are affected because they depend on API access, and if a provider’s new model can’t ship, the platform’s own features that depend on that model can’t ship either.

Low risk: Established model versions. Here’s the important nuance: the government is reviewing new model releases, not existing ones. ChatGPT running on GPT-4o, Claude running on Sonnet 4 — those are already approved and available. What’s being held up is the next version. So if your current workflow works fine on today’s model, it’s not going to break tomorrow.

The risk is future improvements. If you’ve been waiting for GPT 5.6 to handle a specific task better, or for Mythos to finally go live so you can switch — that wait just got a lot longer.

What this means for your subscription decisions

If you’re paying for ChatGPT Plus, Claude Pro, or any AI subscription, you’re paying for access to the latest capabilities. The government approval process means you might be paying for capabilities that can’t ship yet.

This doesn’t mean cancel your subscriptions. Current models are still powerful and useful. But it does mean you should think about what you’re actually getting:

  • ChatGPT Plus ($20/mo): GPT-4o is still available and working. GPT 5.6 is in limited preview. You’re paying for the current model plus whatever comes next — but “next” might be delayed.
  • Claude Pro ($20/mo): Sonnet 4 is available. Mythos is stuck in preview. Same situation — your current model works, but the upgrade path is blocked.
  • Google Gemini: Google’s models haven’t been affected by the government review process yet. This could change, but for now, Gemini is the least likely to face regulatory delays.

If you’re on a tight budget and don’t need cutting-edge features, this might be a good time to consolidate to one subscription instead of two. Pick the one whose current model works best for your use case, and stop paying for a “future” that might be months away.

What I’d actually do right now

1. Test your workflows on multiple models. If you haven’t already, run your key automations through both Claude and GPT. Know which one works better for each task. That way, if one gets delayed, you can switch without scrambling. Here’s how to build model-agnostic automations step by step.

2. Don’t build on unreleased models. If a model is in “limited preview” or “early access,” don’t restructure your business around it. Build on what’s available today. When the new model finally ships broadly, you can upgrade — but don’t plan your roadmap around a release date that nobody can guarantee.

3. Use inference platforms. Services like OpenRouter or Together AI give you access to multiple models through a single API. If one provider’s model gets held up, you can switch to another without changing your code or workflow. This is the model-agnostic approach I covered before, but now it’s not just smart — it’s necessary.

4. Keep your data portable. Your prompts, your templates, your client data — none of that should live exclusively inside one AI platform. Export your ChatGPT conversations, back up your Claude projects, and store your automation templates locally. If you need to switch platforms, you want it to be a config change, not a rebuild.

The bigger picture

I don’t think this is the end of AI innovation. The government approval process is a speed bump, not a roadblock. Models will still get approved, capabilities will still improve, and the tools you use will still get better.

But the pace is changing. The era of “a new model drops every month and everyone gets it instantly” is over for now. We’re entering a period where the gap between “what’s technically possible” and “what you can actually use” might widen. For non-coders, that means being more strategic about which tools you invest in and more flexible about switching when the landscape shifts.

The good news: the tools that are already working today aren’t going to stop working. ChatGPT isn’t going away. Claude isn’t going away. The automation you built last month still runs. The risk is stagnation — being stuck on an older model while newer capabilities sit in regulatory limbo. Build for resilience, not for the bleeding edge.

The bottom line

The government is now a gatekeeper for AI model releases, and both Anthropic and OpenAI are feeling the impact. For non-coders, the practical takeaway is simple: build on what works today, keep your options open, and don’t bet your business on a model that hasn’t shipped yet. If you want help making your AI stack more resilient, check out the AI Tool Advisor — it’ll match you with tools that fit your needs right now, not six months from now.