AMD Goes 2nm, Japan Gets Mythos, and Trump Halts AI Order

Week of June 1, 2026

The White House has paused AI regulation, Japan's largest banks are set to gain access to Anthropic's most powerful model, and AMD has challenged NVIDIA's AI compute dominance. These three stories may seem disconnected, but they all center on the same theme: the battle for control over the next generation of AI infrastructure, from the underlying hardware to the governing regulations.

AMD Fires First at 2nm, Challenging NVIDIA's AI Throne

AMD has commenced production of its 6th Generation EPYC processors, codenamed "Venice," built on TSMC's 2nm process technology. This is the first high-performance computing product to enter production at this node, representing a direct challenge to NVIDIA's dominance in AI compute.

This move is significant because AI training and inference require extreme silicon efficiency. AMD's Venice chips are not only faster but are built on a process that delivers substantially better performance per watt. AMD is already planning a follow-on "Verano" processor to extend its roadmap further.

The timing is strategic. With SK Hynix joining the $1 trillion club this week due to AI memory demand, and NVIDIA CEO Jensen Huang describing Taiwan as the "epicenter" of the AI revolution ahead of Computex, the hardware arms race is accelerating. AMD is signaling that it will not allow NVIDIA to control the AI compute layer uncontested, and the 2nm process provides a competitive advantage.

Japan's Banks Get Claude Mythos — and a Cybersecurity Firewall

Japan's Finance Minister Satsuki Katayama announced that the Japanese government and megabanks MUFG, SMBC, and Mizuho will gain access to Anthropic's Claude Mythos within two weeks. The announcement followed a meeting with US Treasury Secretary Scott Bessent.

However, Katayama also announced a 36-entity public-private working group specifically to address the cybersecurity risks Mythos poses to Japan's financial system. Japan is not blindly opening the gates; it is establishing a containment protocol alongside the access.

This serves as a revealing template for how sovereign governments will handle frontier AI models. Japan is neither banning Mythos nor demanding it be limited. Instead, it is creating a dedicated risk-management structure. Other G7 nations are likely to observe and replicate this model. The critical question is not whether powerful AI will enter national infrastructure, but rather what guardrails are built around it.

Trump Halts AI Executive Order Over China Competition Concerns

President Trump abruptly canceled a scheduled Oval Office signing ceremony for a new AI executive order, stating that the text could undermine America's lead over China. The proposed order emerged from growing pressure within the banking and financial sectors regarding AI cybersecurity risks—particularly those raised by Claude Mythos.

The irony is evident. In the same week Japan moves to securely integrate Mythos into its financial system, the US has halted its own regulatory response due to fears of ceding ground to Beijing. The administration is caught between two instincts: the desire to regulate AI risk and the fear that regulation will slow American innovation relative to China.

This standoff is unlikely to persist. The banking sector's anxiety regarding AI cybersecurity remains, as does the competitive pressure from China. A resolution will likely take the form of a streamlined order that focuses on national security without hindering commercial AI development.

The Bottom Line

Three stories, one signal: AI is entering its infrastructure phase. The hardware race is moving sub-nanometer. Sovereign governments are building bespoke risk frameworks instead of implementing blanket bans. Meanwhile, the White House is discovering that regulating AI without compromising competitiveness is a complex challenge. The next few months will be defined not by model releases, but by who builds the infrastructure and who writes the rules for the data that flows through it.