There’s a question in AI governance that sits unanswered, not because it’s too complex, but because it’s too inconvenient. What happens when the companies developing AI are the same ones drafting the policies meant to regulate it? This isn’t a hypothetical. It’s the current state of play. Look at the major AI ethics councils, the industry summits, the working groups advising governments. The same logos appear on both sides of the table. I’ve read the attendee lists. I’ve cross-referenced the funding disclosures. The overlap isn’t accidental.
This silence on conflict of interest isn’t just an oversight. It’s a decision. By not addressing who gets to write the rules, governments and international bodies are implicitly endorsing a system where the regulated become the regulators. The result is predictable. Policy proposals emerge with built-in carve-outs. Safety standards are framed as voluntary guidelines rather than enforceable mandates. Accountability mechanisms lack teeth because the hands drafting them belong to the same entities that would be held accountable. I’ve pored over the fine print of these frameworks. The language is careful. It suggests responsibility without requiring it.
Consider the pattern. A major AI firm funds a university research center, which then produces a report cited by a government panel—staffed partly by executives from that same firm. The report recommends self-regulation. The panel agrees. The cycle repeats. This isn’t conspiracy. It’s documentation. Public records show these connections in black and white. The EU’s AI Act consultations, the US executive orders on AI safety, the G7’s AI principles—each process includes input from the very companies whose products are under scrutiny. The silence on this dynamic allows it to persist. If no one names the conflict, no one has to resolve it.
What does this mean for the future? It means the rules governing AI will likely prioritize innovation over restraint, profit over precaution. Not because that’s the best path, but because it’s the path drawn by those with the loudest voice in the room. Smaller players—startups, civil society groups, independent researchers—get sidelined. Their input is welcomed in theory, but the final drafts bear the fingerprints of the dominant firms. I’ve seen the comment periods. I’ve tallied who submits and who gets cited. The imbalance isn’t subtle.
This isn’t about villainizing industry. Companies have expertise. Their perspective matters. But when their influence drowns out all others, the result isn’t policy—it’s preference. The silence on this imbalance is a choice to let the current power structure stand. It’s a decision to trust that those with the most to gain from lax rules will somehow advocate for stricter ones. History suggests otherwise. Look at tobacco. Look at finance pre-2008. Self-regulation has a track record, and it’s not a comforting one.
Note for the archive: the question of who writes the rules isn’t abstract. It’s the foundation of what AI will be allowed to become. Ignoring it isn’t neutrality. It’s surrender. The record will show that we saw this coming.
