CASE STUDIES
6 identity-known & 16 anonymous cases published with more to come
COVERAGE
Cases covering 7 AI and tech companies
PARTNER
In collaboration with Blueprint For Free Speech
Read their stories. See how their cases unfolded.
Share the impact they've made with insiders you know.
22 whistleblower cases. Different outcomes. Some stayed at their companies, others founded new organizations. Some catalyzed policy changes, others product and safety improvements.
Anonymous Sources
Case details
- Summaries of 16 anonymous cases from March 2025 – February 2026, where concerns are specifically related to AI safety risks and leadership at frontier AI companies
Case outcomes
- Various; see individual case details
Support received
- If known, provided in individual case details
Lukasz Krupski
Case details
- Product safety risks; lack of safety precautions, autopilot car feature
Case outcome
- Public & regulatory scrutiny; recall of 2M vehicles; won two legal cases – still ongoing
Support received
Where they are now
- Crowdfunding for living and legal expenses
Jack Poulson
Case details
- Product safety risks; authoritarian surveillance through AI tools
Case outcome
- Product termination; public & regulatory scrutiny
Support received
- Collective employee action
Where they are now
Timnit Gebru
Case details
- Suppression of knowledge on AI risks; risks associated with LLMs
Case outcome
- Public & regulatory scrutiny; internal policy and structural changes
Support received
- Public support via social media and petitions; collective employee action
Where they are now
Frances Haugen
Case details
- Product safety risks; social media algorithms amplifying harmful content
Case outcome
- Policy influence (EU Digital Services Act and UK Online Safety Act); public & regulatory scrutiny
Support received
- Legal Support: Whistleblower Aid
Where they are now
Shane Jones
Case details
- Product safety risks; violent and harmful content via AI image-generator
Case outcome
- Changes to product
Support received
- Public support via media and LinkedIn
Where they are now
- Microsoft
What we have learned from the named whistleblower cases
100% of cases resulted in company impact. Majority first reported concerns internally, with most cases then escalated to regulators. Only 2 cases directly influenced new legislation.
View the overall findings.
Last updated: March 2026
Insight 1
Two Core Risk Types Recur
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.
Insight 2
Internal Reporting Is Usually the First Step
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.
Insight 3
External Reporting Follows Internal Failure
In most cases, concerns are later taken to external bodies such as regulators or legislators after internal channels fail to address the issue.
Insight 4
Public Support is Common
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.
Insight 5
Varying Degrees of Impact
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.
Insight 6
Individual Outcomes Vary, But Are Not Universally Negative
While personal consequences differ, many whistleblowers have gone on to found new non-profit organizations or continue working in the field.
Disclaimer: AIWI did not contact any of the companies mentioned in these case studies to corroborate information available in the public domain. This is an evolving resource – more case studies will be added and others may be updated as and when new information arises. No legal advice was sought in the publication of this resource.
What we have learned from the 16 anonymous cases (2025 - 2026)
Last updated: March 2026
Insight 2
Child Safety Failures Are Cross-Company
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.
Insight 4
Disclosure is Being Actively Deterred
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.
Insight 5
Public Claims And Internal Reality Are Diverging
Insight 6
The Same Organisations Appear Repeatedly
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.
Disclaimer: These insights are drawn from publicly reported information. AIWI did not contact any of the companies named to corroborate these accounts.
How we built this resource
AIWI began a search for all known whistleblowing cases at leading AI and tech companies. This led us to identify 70+ whistleblowing cases related to AI safety risks or major tech cases, which helped build awareness among whistleblowers in this industry.
Of the 70+ cases, we focused on those specifically related to Al safety risks or concerns related to the frontier AI companies; six identity-known and sixteen anonymous (22 total) cases have been published to date, with more to follow.
Verified information within the public domain was used to create this resource, including: news articles, company websites, and published documents; US government published sources, e.g., congressional hearing records; and the related whistleblowers’ individual or organizational websites.
The related whistleblowers were also contacted to review, provide feedback, and endorse the final versions of the case studies.
AlWI is grateful for the support of Blueprint for Free Speech in creating this resource, especially to Naomi Colvin, who contributed to the case study of Lukasz Krupski.
Disclaimers: AIWI did not contact any of the leading AI companies mentioned in these case studies to corroborate information available in the public domain. AIWI contacted all whistleblowers to request feedback and any additional thoughts they may want to share about their stories, though a response wasn’t received by all. This is an evolving resource – more case studies will be added and others may be updated as and when new information arises. No legal advice was sought in the publication of this resource. If you have reason to believe any of the information published on this resource is incorrect, please contact AIWI so that it may be corrected.
Find the right support for you
If you'd like AlWI to connect you with pro bono legal counsel with global expertise in whistleblowing and litigation, complemented by AlWI Al expertise.
If you are concerned about costs incurred by good-faith Al whistleblowing, these vetted organizations offer grants and pro bono legal support.
Got questions?
Are you still unsure about your next steps and would like to clarify your support options? Talk to us first and get a free, confidential consultation of your situation.