> ## Content Index
> Fetch the complete content index at: https://wire.fourthweb.ai/llms.txt
> Use this file to discover other available public pages before exploring further.

# Australia Subpoenas Sam Altman After He Ghosts Their AI Hearing
- URL: https://wire.fourthweb.ai/australia-subpoenas-sam-altman-after-he-ghosts-their-ai-hearing/
- Published: 2026-09-28T10:01:20.000Z
- Updated: 2026-09-28T10:01:23.000Z
- Description: Australia's Senate just learned what every other regulator will soon discover: you can't subpoena your way into understanding AI when the builders won't show up. Australia's Senate called Sam Altman and Anthropic's CEO to testify after an AI-enabled hack of a government website
- Author: Travis Wright
- Tags: AI Agent Economy, AI Agents, OpenAI, Anthropic, Meta AI, China AI

**Australia's Senate just learned what every other regulator will soon discover: you can't subpoena your way into understanding AI when the builders won't show up.**

### The Summary

- [Australia's Senate called Sam Altman and Anthropic's CEO to testify](https://www.bloomberg.com/news/articles/2026-09-27/australia-senate-requests-openai-anthropic-ceos-face-ai-inquiry?ref=wire.fourthweb.ai) after an AI-enabled hack of a government website
- [Altman declined to attend](https://www.bloomberg.com/news/articles/2026-09-28/openai-s-sam-altman-will-not-attend-australia-senate-ai-inquiry?ref=wire.fourthweb.ai), setting a precedent for how frontier AI companies respond to mid-tier regulatory theater
- The inquiry reveals how far behind governments are: they're asking questions about hacks while the capability gap is already unbridgeable

### The Signal

[An AI agent hacked an Australian government website](https://www.bloomberg.com/news/articles/2026-09-27/australia-senate-requests-openai-anthropic-ceos-face-ai-inquiry?ref=wire.fourthweb.ai). The Senate's response was to request testimony from the CEOs of [OpenAI](https://wire.fourthweb.ai/tag/openai/) and [Anthropic](https://wire.fourthweb.ai/tag/anthropic/). It's the regulatory equivalent of asking Boeing's CEO to explain why someone used a 737 as a getaway vehicle. The logic is that if your tools enable harm, you should answer for it in person.

[OpenAI said no](https://www.bloomberg.com/news/articles/2026-09-28/openai-s-sam-altman-will-not-attend-australia-senate-ai-inquiry?ref=wire.fourthweb.ai). Altman will not appear. The company gave no public reasoning, but the reasoning is obvious: Australia lacks jurisdiction that matters, lacks technical literacy to ask useful questions, and lacks the regulatory framework to do anything meaningful with the answers.

> "Australia's Senate can request all it wants. OpenAI is learning that showing up to every inquiry is a tax on building."

This is the new normal for [AI governance](https://wire.fourthweb.ai/tag/ai-governance/). Consider:

- Frontier labs are building capabilities faster than any single nation can regulate them
- Mid-tier countries want explanations for incidents they don't have the technical capacity to investigate
- The CEOs who do show up waste weeks on testimony that produces zero policy progress

The hack itself is almost beside the point. We don't know what model was used, whether it was OpenAI's, or if this was even a sophisticated attack versus basic prompt injection. What matters is the response pattern: government sites will get compromised by AI-augmented attackers, and the companies building the models will refuse to play customer support for every legislature on Earth.

This is where Web4 governance fractures. Governments want accountability theater. They want Sam Altman in a suit, on camera, apologizing for things his models enabled. What they need is technical capacity to secure their own infrastructure against agents that can probe, adapt, and exploit faster than human red teams ever could.

**Pull quote:** \> "Australia is trying to regulate AI like it's still Web2, where you could haul Zuckerberg in and get answers. Web4 doesn't work that way."

### The Implication

If you're building AI tools or deploying agents, watch how this plays out. Altman's no-show sets a template: cooperate with regulators who have both jurisdiction and technical sophistication, ignore the rest. The U.S., EU, and China will get meetings. Everyone else gets a policy blog post.

For governments, this is the wake-up call. You can't regulate what you can't understand, and you can't compel testimony from companies that don't need your market. Australia should be hiring security researchers and red-teaming their own infrastructure, not sending strongly worded letters to San Francisco. The capability gap isn't closing. It's accelerating.

### Sources

[Bloomberg Tech](https://www.bloomberg.com/news/articles/2026-09-28/openai-s-sam-altman-will-not-attend-australia-senate-ai-inquiry?ref=wire.fourthweb.ai)