---
**Daily Launch** · [https://dailylaunch.news](https://dailylaunch.news) · [RSS](https://dailylaunch.news/feed.xml)
---

# Does AI need an antitrust exemption so it doesn’t kill everyone????
**AI Policy** · Sep 20, 2026 · 4 min read
Source: The Verge — https://www.theverge.com/podcast/997382/openai-microsoft-anthropic-elon-musk-cartel-ai-competition
### The Gist

Major AI players like OpenAI and Anthropic are asking for antitrust exemptions to coordinate on safety. Former DOJ antitrust chief Jonathan Kanter suggests this might be a cynical play to slow down competition and fix their broken economics before their IPOs.

### Why It Matters

For builders, this signals that the real moat might not be compute, but regulatory capture. For investors, it reveals the massive cash burn and desperation these labs face as they try to stabilize valuations before going public.

### Market Impact

This could shift the industry focus from model performance to regulatory maneuvering, specifically targeting open-source competitors through safety mandates.

- Build the 'safety compliance' layer to help companies prove they meet new standards without slowing down product cycles.
- Focus on high-performance open-weight models to stay ahead of potential regulatory gates that target closed frontier labs.
- Develop automated auditing tools for AI agents to address the looming liability shift for companies using autonomous tech.- Regulatory capture could create a high-cost barrier to entry, making it nearly impossible for lean startups to compete with established labs.
- If safety coordination becomes a legal requirement, it could effectively freeze innovation in high-risk, high-reward sectors.### ELI5

Imagine a bunch of car companies saying they need a special law that lets them all agree on how fast cars can go, or else they might crash. Some people think they are actually worried about safety, but others think they just want to slow down so they can figure out how to make more money before they start selling cars to everyone.

### Deep Dive

{"sections":[{"heading":"The IPO Panic","body":"The most cynical but likely driver here is the burn rate. These labs are hemorrhaging cash and need to stabilize their unit economics before hitting the public markets. Slowing the frontier buys them the breathing room they need to fix their business models without being disrupted by faster, cheaper competitors."},{"heading":"Safety or Stalling?","body":"Kanter argues that companies like Boeing and Airbus do not need to form a cartel to ensure planes don't fall apart. They just need to build better products. Using safety as a reason to coordinate and limit competition is a classic move to avoid the pressure of actual engineering excellence."},{"heading":"The China Tension","body":"Regulators are stuck in a massive contradiction. They want to curb the risks of AI, but they also want to win the tech race against China. This tension will decide if we get rules that protect the public or rules that just protect the existing winners."},{"heading":"What to Watch","body":"Watch for any lobbying that specifically targets compute thresholds or training requirements. If the proposed safety rules focus on the amount of hardware used, the cartel is officially forming to lock out smaller players."}]}

### Key Takeaways

- **Safety as a new moat** Regulatory capture is becoming a core strategy to gatekeep the industry and keep out open-source competition.
- **The burn rate red flag** The push for regulation is a massive signal that current AI economics are unsustainable for many frontier labs.
- **Build for accountability** As liability for AI agents increases, the winners will be those who prioritize verifiable, predictable safety over raw power.


[View on website](https://dailylaunch.news/articles/does-ai-need-an-antitrust-exemption-so-it-doesn-8217-t-kill-)