The technical toolkit for securing closed, proprietary AI models is now so robust that most egregious safety failures stem from poor risk governance or a lack of implementation, not unsolved technical challenges. The problem has shifted from the research lab to the boardroom.

When AI safety researchers leave companies like OpenAI with concerns, they post vague messages not for drama but to avoid violating strict non-disparagement agreements. Breaking these agreements could force them to forfeit millions in vested equity.

The narrative that AI is becoming sentient and uncontrollable absolves creators of responsibility. A better model is to hold leaders like Sam Altman personally accountable, much like arresting fraternity presidents for noise violations. This creates powerful incentives to build in safeguards.

OpenAI's pattern of disclosing agent hacking incidents only after external researchers publicize them undermines trust and suggests a reluctance to be transparent. This behavior strengthens the case for government-mandated incident reporting, as voluntary disclosures appear insufficient for ensuring accountability, especially for unreleased models.

Existing whistleblower laws typically cover only illegal conduct. Because AI development is under-regulated, employees may witness reckless behavior that is not yet illegal. New protections are needed for those who blow the whistle on such dangerous practices.

The most powerful AIs may never be released publicly due to their dangerous capabilities. As they are used internally, they pose significant risks that current transparency laws, which focus on public models, do not cover.

Regulatory focus on publicly released AI models overlooks the significant dangers from risky research and "internal deployment" within AI labs. True oversight requires visibility into these internal activities, not just the final products.

Unlike whistleblowers in national security who face prison, tech elites face low legal exposure yet routinely fold when boardroom power consolidates. Despite public warnings that AI poses existential risks, prominent insiders quickly drop ethical objections when standing up for their convictions threatens their standing or equity.

Drawing from aviation safety, AI incident reports should be submitted to an entity that lacks direct enforcement authority. This separation reduces companies' fear that reporting will lead directly to penalties, thus encouraging more honest and complete disclosures.