The launch of the Chinese AI startup Moonshot's open-source model, Kimi K3, has sent shockwaves through Washington policy circles, thrusting NVIDIA CEO Jensen Huang to the forefront of the debate.
According to reports, Huang traveled to Washington on Tuesday for a series of high-level meetings with lawmakers and Commerce Secretary Howard Lutnick. His central argument is that open-weight models are a strategic necessity for the security and competitiveness of the US AI industry.
Following meetings with Democratic Senators Mark Warner and Adam Schiff, Huang stated, "For the US industry, we need open weights for safety, we need open weights for security. This is vital for the entire industry's dynamism." This marks the latest escalation in a public lobbying campaign that has been building since the release of Kimi K3.
The Context of the Defense
The backdrop to this "defense campaign" is the public release of the Kimi K3 model on Monday, which is now available for open download. Its performance is reportedly comparable to offerings from Anthropic and OpenAI, reigniting policy debates in the US over potential restrictions on open-weight models.
Analysts suggest that the outcome of these policy discussions will directly impact the commercial prospects of AI hardware vendors like NVIDIA. Open-source models, being cheaper to deploy and more widely used, can drive a significant increase in revenue for the underlying computing power.
Joint Pressure: A Collective Industry Front
A spokesperson for NVIDIA stated that Huang's trip was aimed at discussing the company's efforts to support "US AI leadership," including its investments in the open-source ecosystem.
The timing of these meetings is critical, as the Commerce Department is finalizing its security review framework for advanced AI models. This means Huang's lobbying is directly targeting a key policy-making window, potentially shaping the upcoming regulatory framework.
Huang's Washington trip is not a solo effort but part of a broader, organized industry campaign. Last Friday, he joined Microsoft CEO Satya Nadella and dozens of other tech executives in a letter warning policymakers against restricting the open-weight AI development path.
The letter described open-weight models as "a foundational component of the ecosystem because they make advanced AI more accessible, adaptable, and more widely available." Huang used the letter as his first post on his newly launched personal X account.
On Monday, NVIDIA again pressed the issue, releasing a statement alongside Microsoft, Palantir Technologies, and Salesforce. This statement positioned open-weight models as a critical tool for strengthening cybersecurity, using strong language against blanket restrictions.
Within a matter of days, two joint letters and a series of targeted meetings on Capitol Hill have escalated the intensity of Huang's policy campaign.
Reports indicate that Huang's lobbying efforts are already showing some results. Senator Mark Warner, a member of the Senate Intelligence Committee, said after meeting with Huang that the CEO had made a compelling case for open-weight systems.
"Now that we see Chinese models like Kimi emerging, and more US companies are also moving to open-source, I'm not sure this is a genie that can be put back in the bottle," Warner commented.
However, not all lawmakers are receptive to Huang's visit. Democratic Senator Elizabeth Warren, who had previously invited Huang to testify publicly before the Banking Committee, expressed disappointment when he declined.
"He can't find an hour to answer questions publicly before Congress?" Warren said in a statement, highlighting the pressure Huang still faces from some legislators in Washington.
The Business Logic: Why an Open-Source Wave Benefits NVIDIA Hardware
According to reports, Huang's push for open-weight models is also underpinned by clear commercial logic. While US AI developers like OpenAI offer some open-weight models, their core business often revolves around selling closed or proprietary systems.
In contrast, Chinese companies like DeepSeek have already released a wave of cheaper open-source models. Analysts believe that the proliferation of open-source models lowers the barrier to entry and enables broader deployment, all of which ultimately requires significant computing power.
As the dominant supplier of AI chips, NVIDIA stands to benefit from a massive increase in hardware demand as the open-source ecosystem expands. In other words, the success of open-source models is commercially aligned with the interests of NVIDIA.
Comments