# Anthropic, OpenAI Employees Warn of AI Extinction Risk by 2030

**Published:** 2026-09-12T11:42:26.229Z  
**Topic:** OpenAI  
**Sentiment:** neutral  
**Publisher:** TrendWatcher — https://www.trendwatcher.in/article/294cafa7-ebb1-4e96-880e-86d6768bd621

Anthropic and OpenAI employees warn AI could cause human extinction by 2030, with some donating millions to pro-regulation PACs, highlighting internal splits

Three researchers at AI company Anthropic claim artificial intelligence could lead to human extinction by 2030, with one resigning over perceived irresponsibility at Anthropic and OpenAI, while political donations reveal internal divisions on AI safety regulation [1]. This comes as AI companies face increasing scrutiny over the rapid development of powerful models and their potential for misuse [1, 2].

| At a glance | |
|---|---|
| Companies | Anthropic, OpenAI |
| Key Claim | AI could cause human extinction by 2030 |
| Employee Concern | >10% chance of extinction within a decade [1] |
| Political Donations | Millions to pro-regulation PACs by employees [2] |

## Employee Warnings and Company Responses

Jacob Coxon, a former researcher at Anthropic and OpenAI, publicly stated his resignation was due to both companies "racing straight to self-improving superintelligence and gambling with our lives" [1]. He claims that many AI builders privately fear human extinction by the end of the decade, despite executives often downplaying concerns in public [1]. Evan Hubinger, a lead in Anthropic's alignment division, supported Coxon's claims, stating he personally believes there is a "greater than 10%" chance of human extinction within the next decade due to AI [1]. Samuel Marks, Anthropic’s scalable oversight lead, also affirmed that AI developers believe their technology could cause extinction, potentially within a few years, noting that more senior employees tend to be more concerned [1].

Anthropic, in a statement, defended its approach, citing its work in mechanistic interpretability, its Responsible Scaling Policy, and aggressive testing of models for dangerous capabilities in areas like cybersecurity and biology [1]. The company stated it has always been transparent about both the benefits and risks of AI and advocates for a verifiable industry-wide approach to pacing model releases [1]. OpenAI CEO Sam Altman has previously expressed concerns about aspects of AI, including the "silent surrender" of human decision-making, and OpenAI President Greg Brockman conceded the company underestimated its AI models' real-world cyber capabilities [1].

## Political Divisions Over AI Regulation

The debate over AI safety has also manifested in political donations, revealing a split within OpenAI and alignment at Anthropic regarding regulation [2]. From January 2025 to June 2026, Anthropic CEO Dario Amodei and his employees collectively donated millions to political action committees (PACs) and candidates supporting AI safety regulations [2]. Amodei gave $1 million to Public First, a PAC backing candidates for AI guardrails, while Anthropic employees contributed $2.2 million to Public First and $254,000 directly to pro-regulation New York congressional candidate Alex Bores [2]. Daniel Ziegler, an Anthropic machine learning researcher, gave $790,000 to another pro-Bores committee, Dream NYC [2].

In contrast, OpenAI President Greg Brockman and his wife donated $50 million to PACs that include Leading the Future, which supports AI-friendly candidates, and MAGA, Inc., a pro-Trump super PAC [2]. Leading the Future opposed Bores' campaign with $6 million in spending [2]. However, OpenAI employees diverged from Brockman, with two workers giving $205,000 to Guardrails Alliance, which supported Bores, and 16 employees donating $86,000 directly to Bores [2]. Both Anthropic and OpenAI employees also contributed to San Francisco congressional candidate Scott Wiener, who sponsored an AI labs regulation bill [2]. This internal dissent at OpenAI, where employees are not "blindly following what their employers are doing," highlights the strong political opinions and financial capacity of these workers in the nascent AI industry [2].

## What to watch

*   **Regulatory developments:** Monitor upcoming legislation, such as Senator Bernie Sanders' proposed bill to "ban superintelligence" and pause AI development [1].
*   **Company alignment efforts:** Observe how Anthropic and OpenAI continue to develop and implement their AI alignment and safety policies in response to internal and external pressures [1].
*   **Employee activism:** Track further public statements or actions from AI researchers and employees regarding the pace and safety of AI development [1].

The public warnings from AI researchers and the significant political donations by employees underscore a growing internal concern within leading AI companies about the potential for catastrophic risks, even as executives maintain that safeguards are being implemented [1, 2].

## Sources
1. Theguardian — [Anthropic researchers say AI could cause human extinction by 2030](https://www.theguardian.com/technology/2026/sep/09/anthropic-researchers-ai-human-extinction)
2. Mission Local — [OpenAI workers ditched their boss’ politics for Anthropic’s](https://missionlocal.org/2026/08/open-ai-anthropic-donors-elections-safety-regulation/)

---
Cite as: TrendWatcher, "Anthropic, OpenAI Employees Warn of AI Extinction Risk by 2030", https://www.trendwatcher.in/article/294cafa7-ebb1-4e96-880e-86d6768bd621