Common Sense Media, the nonprofit that rates everything from movies to apps for parents, just made the boldest move in AI child safety to date. According to Politico, the organization is soliciting tens of millions of dollars from top AI companies to launch an independent institute that would assess the risks AI technologies pose to children.

The ask: $100 million from each major AI company. The pitch: let an independent body with actual credibility evaluate your products before kids use them. The subtext: we know you will not do this yourselves, because you already proved it.

The timing here is devastating for OpenAI. Just days ago, we reported that OpenAI secretly funded a grassroots-looking child safety coalition to write its own safety standards. Nonprofits were furious. The coalition looked independent but was bankrolled by the company whose products it was supposed to evaluate. Classic tech playbook: if you do not like the referee, become the referee.

Now Common Sense Media is offering the alternative. And unlike OpenAI's astroturf operation, Common Sense has decades of credibility with parents, schools, and policymakers. When they rate something, people listen. When they say a product is unsafe for kids, school districts act on it.

Why This Matters More Than Another Safety Pledge

The AI industry has made approximately 847 voluntary safety commitments over the past three years. None of them are enforceable. None of them are independently verified. They exist to absorb political pressure without changing behavior.

What Common Sense Media is proposing is structurally different. An independent institute, funded by industry but governed independently, that publishes assessments companies cannot control. Think Underwriters Laboratories for AI: you do not get to sell the product until someone outside your building says it is safe.

The question is whether AI companies will actually pay for oversight they cannot influence. OpenAI's answer, based on recent behavior, appears to be: only if we are the ones writing the rules. Anthropic, which has built its entire brand on safety-first principles, faces a different calculation. Saying no to this would undermine the one thing that differentiates them from the competition.

The Political Window

This proposal lands while Congress is actively debating the TRUMP AMERICA AI Act, which includes its own child safety provisions. The White House framework lists protecting children as its first pillar. Both parties agree AI needs to be safer for kids. Nobody agrees on who should enforce it.

Common Sense Media is positioning itself as the answer to that question. Not the government, which moves too slowly. Not the companies, which have a conflict of interest. An independent body with the credibility to make assessments that actually matter.

If OpenAI says yes, they are admitting their own safety efforts were insufficient. If they say no, they are confirming what everyone already suspects: safety is a marketing strategy, not a company value. Either way, Common Sense Media just put them in a corner they cannot talk their way out of.