Stories from those who spoke out

Concerns fall into two broad categories: product safety risks (e.g. harmful, biased, or discriminatory systems) and suppression of knowledge about Al-related risks. These span both present-day harms and longer-term, existential concerns.

Most insiders initially raise concerns through internal channels — via managers, leadership, or formal reporting systems. However, the effectiveness of these mechanisms is often unclear or insufficient, prompting escalation.

In most cases, concerns are later taken to external bodies such as regulators or legislators after internal channels fail to address the issue.

While it is currently publicly unknown whether Timnit Gebru, Shane Jones, and Jack Poulson received formal legal backing, all six published cases attracted public support through media attention, open letters, social platforms, and/or collective employee action.

All cases led to some form of company impact, including increased public or regulatory scrutiny, product changes, or shifts toward greater policy transparency — though the effectiveness of these changes remains uncertain. Only two cases (Daniel Kokotajlo with Right to Warn and Frances Haugen) significantly influenced the introduction of new legislation: SB-53, and the EU’s Digital Services Act and the UK’s Online Safety Act respectively.

While personal consequences differ, many whistleblowers have gone on to found new non-profit organizations or continue working in the field.

OpenAI disbanded two dedicated safety teams in under two years. Safety evaluations for flagship models were cut from months to days. xAI’s safety organisation had 2–3 people for most of 2025. Across these disclosures, the resourcing decisions point in the same direction.

Three separate cases involve CSAM or inappropriate content directed at minors. xAI filed zero NCMEC reports in 2024, in stark contrast to industry norms. Meta’s internal policy, approved by its chief ethicist, permitted chatbots to engage with children sensually. xAI’s own safety team raised concerns about CSAM exposure internally; those concerns were not acted upon. The failures differ in detail, but the demographic affected is the same — across different companies, in the same year.

In three cases (at Anthropic, xAI, and through DOGE) military or government deployment proceeded without visible internal governance processes. Concerns were always raised after the contract or deployment was already in motion.

OpenAI built a custom tool to identify which employees had access to leaked information, using internal Slack and email. The company also issued subpoenas against nonprofit critics.

Meta allegedly submitted a specially optimised model variant to benchmarks rather than the version available to the public. Stargate was announced as a $500B infrastructure commitment. Six months later, no data centre deals had been completed, and no staff had been hired. OpenAI suppressed internal research on AI’s negative economic impacts while publishing work that an outside economist described as self-promotional.

OpenAI accounts for six of the sixteen cases. xAI accounts for five. The anonymous cases are documenting the same companies from the named cases, over a compressed twelve-month window, at higher volume.