The guardrails finally showed up
Regulation reached the AI tools this week. That is healthy, and the question is who each rule protects.
Mariano De Vitto · June 2026
This was the week the rules caught up with the tools. OpenAI shipped its most powerful model, GPT-5.6, but not to the public. It went to about 20 pre-approved organizations under a federal review framework created weeks earlier. Google confirmed that buying citations in AI Mode and AI Overviews now counts as spam. Meta started charging location fees tied to digital service taxes and faced fresh questions about training ads on user data. After two years of ship first and explain later, the guardrails finally showed up.
This is good news, mostly. A market without rules is not free, it is just unaccountable. When the most capable model on earth can ship to anyone overnight, the people downstream, including every marketer who builds on it, inherit risks they never agreed to. A review step before wide release, clear lines on what counts as manipulation, honest disclosure on ads: these are the conditions that let a serious industry plan. The marketer who wanted regulation to stay away was really asking to keep operating on ground that could shift without warning. Rules are what make the ground stable enough to build on.
But rules are not neutral
Here is the part to watch. Every rule protects someone, and not always the person you would assume. A government review framework for AI models sounds like public safety, and it may be, but it also decides who gets early access to the most powerful tools, which is a competitive advantage handed out by policy rather than earned in the market. Google deciding what counts as citation spam protects search quality, and it also lets Google set the terms of visibility in a system it owns end to end. Disclosure rules protect consumers, and they also shift cost and liability onto advertisers. None of this is a reason to oppose the rules. It is a reason to read them closely and ask the question that matters: who does this protect, and who pays for it.
What a marketer should actually do
Three moves. First, treat access as a variable, not a constant. If the strongest model ships to 20 companies first, your plan cannot assume you are one of them, so build for the capability you can reliably reach and upgrade when access widens. Second, get your own house in order before a regulator or a platform does it for you: honest disclosure, clean data consent, a human who signs off on claims. The brands that already operate this way will find the new rules easy. The ones cutting corners will find them expensive. Third, keep a record. As agents take over more of the buying and the messaging, the ability to show what happened and why becomes the difference between a defensible decision and an unexplainable one.
The through line of this whole year, from the Cannes jury to the agent layer to this week's guardrails, is the same. The technology keeps getting more capable and more autonomous, and the human job keeps narrowing to the parts that actually require judgment: what is true, what is fair, who is accountable. Rules do not remove that job. They formalize it. The marketer who treats the guardrails as a burden will spend the year fighting them. The one who treats them as the new floor will build on top.
The signal: Rules are arriving, and that is healthy. Read who each one protects, get your disclosure and data house in order, and keep a human accountable for what the machine does in your name.
The Signal Brief · Mariano De Vitto — Head of Marketing, Barcelona