AI Insiders at Major Labs Raise Public Alarms About Their Employers

What Happened

A growing number of researchers and employees at leading artificial intelligence companies, including OpenAI, Anthropic, and Google DeepMind, have spoken out publicly about concerns at their organizations, according to a report from Fast Company. The disclosures span a range of issues including copyright infringement, military applications of AI systems, and broader safety concerns.

Who Is Speaking Out

The individuals raising concerns work or have worked inside some of the most prominent AI laboratories in the world. OpenAI is the San Francisco-based company behind ChatGPT and GPT-4, and one of the most heavily funded AI firms globally. Anthropic, also based in San Francisco, is the company behind the Claude family of AI models. Google DeepMind is the AI research division of Alphabet, Google's parent company.

The Fast Company report describes the list of insiders speaking out as continuing to grow, with employees and researchers across multiple organizations choosing to go public with concerns rather than raising them solely through internal channels.

What the Concerns Cover

According to the report, the issues being raised by insiders include alleged copyright infringement related to AI training data, the use of AI systems in military contexts, and unspecified concerns about the pace and direction of AI development. The report does not detail specific internal policies or incidents at each company by name, but indicates the concerns are varied and come from multiple levels within these organizations.

Separately, former Anthropic researcher Jacob Coxon stated in a podcast interview, reported by BigGo Finance, that AI technology has become powerful enough to be considered a superweapon. Coxon warned that conditions in 2027 could bring significant instability. Coxon is identified as a former researcher at Anthropic, not a current employee.

Background

Internal dissent at AI companies is not new. In 2024, a group of current and former OpenAI employees published an open letter calling for stronger protections for AI safety whistleblowers. Several employees resigned from OpenAI around the same period citing safety-related disagreements. Anthropic itself was founded in 2021 by former OpenAI employees, including Dario Amodei and Daniela Amodei, who cited concerns about AI safety practices at their prior employer.

The pattern of insider disclosures has prompted discussion in policy circles about whether existing legal protections for technology whistleblowers are sufficient. In the United States, no federal legislation specifically governs AI whistleblowing, though broader protections under securities law and labor law may apply in some circumstances.

All three companies named in the reports, OpenAI, Anthropic, and Google DeepMind, have published safety frameworks and acceptable use policies governing how their models can be deployed. OpenAI and Anthropic have each described safety as a foundational commitment in public statements and regulatory filings.

What It Means in Practice

Public statements from employees and former employees can carry legal and reputational implications for the companies involved. At minimum, they place on the public record concerns that might otherwise remain internal. The specific claims cited in the Fast Company report, particularly around copyright and military use, touch on areas already subject to active litigation and regulatory scrutiny.

Copyright lawsuits involving AI training data are currently moving through federal courts in the United States, with cases brought by publishers, authors, and visual artists against multiple AI companies. Military AI procurement and deployment is also under review by the U.S. Department of Defense, which has published ethical guidelines for AI use in defense contexts.

None of the three companies had issued formal public responses to the Fast Company report at the time of this writing.

What Comes Next

Congressional committees in the United States with oversight of technology and national security have previously scheduled hearings on AI safety and accountability, and additional testimony from industry insiders is expected as those proceedings continue.