# Musk's AI flip-flops reveal the safety debate's real stakes

> Elon Musk agrees with rivals on slowing AI, then contradicts himself by fighting regulation and selling weapons-grade models to the military.
- Title: Musk's AI flip-flops reveal the safety debate's real stakes
- Summary: Elon Musk agrees with rivals on slowing AI, then contradicts himself by fighting regulation and selling weapons-grade models to the military. Musk's shifting…
- Keywords: elon musk, ai safety, openai, anthropic, xai, spacex, sam altman, dario amodei, regulation, self-regulation, ai development pace, grok
- Source: CNBC — https://cnbc.com/2026/09/18/after-decade-of-clashes-in-ai-elon-musk-forging-strange-alliances.html
- Author: Lora Kolodny
- Published: 2026-09-18T19:15:24+00:00
- Read time: 2 min
- Topics: elon musk, ai safety, openai, anthropic, xai, spacex, sam altman, dario amodei, regulation, self-regulation, ai development pace, grok
## Musk's wild week of AI reversals
In one week, Musk agreed with rivals on slowing AI development, then was contradicted by Trump and Nvidia's Jensen Huang who said companies need to move faster.
## The middle path Musk proposes
Musk urged companies to test each other's models for safety issues before release, rather than letting government impose heavy-handed regulation that's hard to reverse.

> You can always escalate the amount of regulatory oversight, but it is very difficult to reduce it.
## Behind the scenes with Trump
The Wall Street Journal reported Musk joined Nvidia's Huang and Meta's Zuckerberg in advising Trump to oppose an industry-funded regulator.
## A decade of AI conflicts
Musk funded DeepMind in 2013, co-founded OpenAI in 2015, then clashed with Sam Altman over control and profit structure, eventually suing him in 2024.
## The fear-mongering pattern
Since 2014, Musk has warned of AI superintelligence threats like Terminator scenarios while simultaneously building his own AI businesses and resisting regulation.

> There have been movies about this, you know, like '[The] Terminator.' There are some scary outcomes.
## The demon summoning speech
In late 2014, Musk told MIT students AI development was like summoning the demon, then donated $10 million to the Future of Life Institute to raise awareness of AI risks.
## OpenAI's nonprofit promise broken
Musk co-founded OpenAI in 2015 as a nonprofit to benefit humanity, but clashed with Altman in 2018 over control and profit structure, eventually leaving the board.
## ChatGPT left Musk behind
When OpenAI released ChatGPT in November 2022, kicking off the generative AI boom, Musk was largely sidelined after his 2018 departure.
## Musk launches xAI in secret
In March 2023, Musk signed another pause AI letter, then quietly incorporated xAI in Nevada without telling Tesla investors until July.
## The Anthropic flip-flop
In February, Musk called Anthropic misanthropic and evil. By May, after Anthropic agreed to pay SpaceX $1.25 billion monthly for computing, Musk called them AI leaders.

> I was clearly wrong about Anthropic. They are obviously currently the leader in AI.
## SpaceX profits from military AI
SpaceX sold Grok models to the U.S. military, enabling forces to deploy over 2,000 munitions to 2,000 targets within 96 hours, according to DOJ filings.
## Fighting state AI regulations
xAI sued to stop Minnesota's ban on Grok's nudify features and California's transparency requirements for training data, even as Musk calls for safety.
## Memphis data center pollution
Musk built xAI data centers in Memphis while facing litigation from residents suffering air and noise pollution from the operations.
## Self-regulation won't work
Tyler Whitmer of the nonprofit Legal Advocates for Safe Science and Technology told CNBC that companies cannot police themselves. Real government regulation with third-party evaluators is needed.

> We need real regulation that creates a level playing field for companies but also protects the public, and the companies just can't do that themselves.
## The contradiction at the core
Researchers documented how Musk and other AI leaders claim to safeguard humanity while creating unsafe products, centralizing power, and evading accountability.
## Key takeaway

Musk's shifting positions on AI safety expose a fundamental tension: industry leaders talk about existential risks while building profitable businesses that resist oversight.