Democratic congresswoman Lori Trahan has criticised the Trump administration’s decision to temporarily bar Australian government security agencies and Australian citizens working on AI in the US from working with the powerful Claude Mythos AI tool, saying the United States should foster more open collaboration with its Five Eyes allies.
Trahan, a Massachusetts congresswoman and member of the House Committee on Energy and Commerce, said the technology should not have been withheld from Five Eyes intelligence partners. Speaking at the Takeoff AI conference in Washington DC, she called for a unified “Western stack” of software tools. “Innovation and safety are not at odds with each other, and I think we should just remind ourselves that when we set those standards high, we bring the rest of the world with us. But that's also incumbent on our collaboration with our allies,” she said.
Anthropic released Claude Mythos in April. The model is particularly skillful at finding and exploiting vulnerabilities in IT systems. The Trump administration initially barred Australian government security agencies and Australian citizens working on AI in the US, along with other foreign nationals, from working with it, before partly winding back the restrictions after a couple of months. Anthropic has heavily restricted access to financial institutions, major tech companies and cloud providers.
Trahan told this masthead: “We rely on those relationships with our allies, and so I think that there's a collaboration that we need to strike.”
David Wroe, director of AI security organisation Good Ancestors, welcomed Trahan's comments, describing Mythos as a tool with the power to “carry out any cyber activity you can imagine”. He said it was very much in Australia's interest to have infrastructure built in Australia. Anthropic is examining Australia as one of a handful of countries for a second location for AI training data centres.
The Australian government recently announced an AI Safety Institute in response to a push for greater regulation of artificial intelligence domestically, something which has been anathema to the Trump administration.
Analysis
Why This Matters
- The debate over Claude Mythos access highlights the tension between US efforts to control the most powerful AI tools and the need for close cooperation with its intelligence allies.
- Australia's potential as a host for Anthropic's training data centres positions the country at the centre of the global AI infrastructure buildout.
- The push by Democrats in Congress for a "Western stack" of AI tools signals a geopolitical strategy that seeks to bind allied countries together around shared technology standards.
Background
In April, Anthropic released Claude Mythos, an AI model designed to find and exploit IT vulnerabilities. The Trump administration blocked Australian agencies and citizens from accessing it, a restriction that was later partly reversed. Trahan is helping lead a push for AI regulation in the US and believes access for Five Eyes partners should be prioritised.
Key Perspectives
Lori Trahan (Democratic Congresswoman, Massachusetts): Says the US should prioritise collaboration with Five Eyes partners and that withholding the tool was a mistake. She argues innovation and safety are compatible.
David Wroe (Director, Good Ancestors): Welcomed Trahan's position, saying Australia wants to be first in line for access to advanced US technology and that building AI infrastructure in Australia is in the country's interest.
What to Watch
- Whether the Democrats' new majority in Congress can enact legislation changing how AI tools are shared with allies.
- Anthropic's decision on where to build its next training data centre, with Australia under active consideration.
- Any signal from the Trump administration on whether it will soften its stance on AI regulation and international collaboration.
Checked
12 claims in this story checked against 1 source: 11 supported.
Show each claim
- SupportedLori Trahan says Australia should get priority access to the powerful Claude Mythos AI tool.Sydney Morning Herald - Latest News: “Australia should be given priority access to the most capable artificial intelligence tools developed in the US and the Trump administration was wrong to temporarily withhold access to the powerful Claude Mythos tool. That’s the view of Lori Trahan”
- SupportedTrahan criticised the Trump administration’s decision on Claude Mythos and called for closer Five Eyes technology collaboration.Sydney Morning Herald - Latest News: “She believes that as part of the push for better “rules of the road” to ensure the technology is properly controlled, the US should also foster more open collaboration with its Five Eyes intelligence partners, particularly Australia and the UK.”
- SupportedThe Trump administration initially barred Australian government security agencies and Australian citizens working on AI in the US from using Mythos, then partly wound back the restrictions after a couple of months.Sydney Morning Herald - Latest News: “Australian government security agencies and even Australian citizens working on AI in the US – along with all other foreign nationals – were initially barred from working with Mythos after a Trump administration edict, which was partly wound back after a couple of months.”
- SupportedTrahan said the technology should not have been withheld from Five Eyes intelligence partners.Sydney Morning Herald - Latest News: “Trahan, who is a member of Congress’ House Committee on Energy and Commerce, told this masthead the technology should not have been withheld from Five Eyes intelligence partners.”
- SupportedAt the Takeoff AI conference in Washington DC, Trahan said: “Innovation and safety are not at odds with each other, and I think we should just remind ourselves that when we set those standards high, we bring the rest of the world with us. But that's also incumbent on our collaboration with our allies”.Sydney Morning Herald - Latest News: “Speaking at the Takeoff AI conference in Washington DC, Trahan said “there’s no question that AI was on the ballot” during the recent mid-term elections”
- Partly supportedAt the Takeoff AI conference, Trahan called for a unified “Western stack” of software tools.Sydney Morning Herald - Latest News: ““We rely on those relationships with our allies, and so I think that there’s a collaboration that we need to strike. In fact, we should have a Western stack [a collection of software tools] that is unified and fostering collaboration,” she said.”
- SupportedAnthropic released Claude Mythos in April.Sydney Morning Herald - Latest News: “In April, Anthropic released its Claude Mythos AI model, which is particularly skillful at finding and exploiting vulnerabilities in IT systems.”
- SupportedAnthropic has heavily restricted Mythos access to financial institutions, major tech companies and cloud providers.Sydney Morning Herald - Latest News: “Anthropic, which makes the Claude AI chatbot and is on the verge of a record $USD1.4 trillion initial public offer, has heavily restricted access to Mythos to financial institutions, major tech companies and cloud services providers like Amazon.”
- SupportedTrahan told the masthead: “We rely on those relationships with our allies, and so I think that there's a collaboration that we need to strike.”Sydney Morning Herald - Latest News: ““We rely on those relationships with our allies, and so I think that there’s a collaboration that we need to strike. In fact, we should have a Western stack [a collection of software tools] that is unified and fostering collaboration,” she said.”
- SupportedDavid Wroe said Mythos had the power to “carry out any cyber activity you can imagine”.Sydney Morning Herald - Latest News: “David Wroe, director of AI security the organisation Good Ancestors, welcomed Trahan’s comments and said Mythos had the power to “carry out any cyber activity you can imagine”.”
- SupportedAnthropic is examining Australia as one of a handful of countries for a second location for AI training data centres.Sydney Morning Herald - Latest News: “Australia is one of a handful of countries being examined by Anthropic as a second location to set up AI training data centres”
- SupportedThe Australian government recently announced an AI Safety Institute in response to a push for greater domestic AI regulation, which has been anathema to the Trump administration.Sydney Morning Herald - Latest News: “The Australian government recently announced an AI Safety Institute in response to a push for greater regulation of artificial intelligence domestically, something which has been anathema to the Trump administration.”