⚡NIMITZ TECH NEWS FLASH⚡

“Winning the AI Race: Strengthening U.S. Capabilities in Computing and Innovation”

Senate Committee on Commerce, Science, and Transportation

May 8, 2025 (recording linked here)

HEARING INFORMATION

Witnesses and Written Testimony (Linked):

• Mr. Sam Altman: Co-Founder and Chief Executive Officer, OpenAI

• Dr. Lisa Su: Chief Executive Officer and Chair, Advanced Micro Devices (AMD)

• Mr. Michael Intrator: Co-Founder and Chief Executive Officer, CoreWeave

• Mr. Brad Smith: Vice Chair and President, Microsoft Corporation
  

HEARING HIGHLIGHTS

IN THEIR WORDS

SUMMARY OF OPENING STATEMENTS FROM THE COMMITTEE AND SUBCOMMITTEE

• Committee Chair Cruz opened the hearing by declaring that artificial intelligence has brought the world to a critical inflection point, comparing its impact to that of the internet. He stressed that U.S. leadership in AI is vital for both economic and national security, warning that China is aggressively pursuing dominance by 2030. Cruz criticized the Biden administration’s AI regulatory approach, likening it to Europe’s heavy-handed style, which he blamed for stifling innovation. He advocated instead for a light-touch regulatory model modeled on the early internet era, announcing a forthcoming bill to create an AI regulatory sandbox to accelerate U.S. innovation and global competitiveness.

• Committee Ranking Member Cantwell emphasized the importance of an "open AI" approach to defeat China’s closed systems. She highlighted the need for U.S. leadership in computing power, algorithms, and data, and praised the CHIPS and Science Act for strengthening domestic supply chains. Cantwell called for smart planning around energy use and data centers, applauded public-private efforts like Microsoft’s fusion energy deal, and stressed the urgent need to train more electricians. She warned that delay in setting U.S.-led standards could lead to another Huawei-like scenario and urged bipartisan cooperation to enact pending legislation that supports American innovation and global AI leadership.

SUMMARY OF WITNESS STATEMENT

• Mr. Altman reflected on the rapid evolution of ChatGPT from a curiosity two years ago to a tool used by over 500 million people weekly, significantly increasing productivity across fields like science, medicine, and education. He emphasized the importance of U.S. leadership in AI, likening its potential impact to the internet and stressing the need for investment in infrastructure to maintain momentum. Sharing a personal story from his upbringing during the internet boom, Altman attributed his success to America’s culture of innovation and warned against adopting overly restrictive policies that might hinder future revolutions. He closed with an appeal to preserve America’s role as the birthplace of transformative technologies by supporting the next generation of innovators.

• Dr. Su highlighted AMD’s critical role in the AI ecosystem, powering defense systems, communications, and the world’s fastest supercomputers through longstanding public-private partnerships. She emphasized that while the U.S. currently leads in AI, global competition is fierce and leadership is not guaranteed. To maintain an edge, she proposed five priorities: accelerate computing infrastructure, support open ecosystems, strengthen domestic manufacturing, invest in talent, and balance national security with global tech adoption. Dr. Su concluded with a personal anecdote about growing up learning on early computers, reinforcing that the U.S. remains the best place for computing innovation and must stay that way.

• Mr. Intrator described CoreWeave’s transformation from a garage startup into a leader in AI infrastructure, now operating over 30 data centers and managing 250,000 GPUs with $1.9 billion in revenue. He stressed that modern AI workloads require infrastructure far beyond traditional cloud capabilities, demanding vast computing power, advanced cooling, and high-speed networking. He outlined four key policy needs: stable investment frameworks, expanded energy infrastructure, global market access, and strong public-private partnerships for workforce development. Intrator concluded that the U.S. is at a pivotal moment, and coordinated action across sectors is necessary to secure AI leadership and unlock economic and societal benefits.

• Mr. Smith emphasized that U.S. AI leadership depends on a collaborative ecosystem involving infrastructure, platforms, and applications, and noted Microsoft’s partnerships with other witnesses. He identified three national priorities: accelerating innovation through infrastructure and research, ensuring widespread adoption via workforce skilling, and sustaining global leadership through responsible exports and international trust. Smith argued that AI should be built to empower—not replace—people, and emphasized the need for investment in education and upskilling to ensure that all Americans can benefit from AI advancements. He ended by urging policymakers to create conditions for young people to thrive with technology, echoing the origin stories of industry leaders.

SUMMARY OF KEY Q&A

• Sen. Sheehy asked each witness to describe, in five words or less, what they needed from the U.S. government to win the AI race. Mr. Smith answered “more electricians” and “broader AI education.” Mr. Intrator called for faster infrastructure development. Dr. Su said policy should speed innovation. Mr. Altman replied “supply chain–sensible policy.”

  Sen. Sheehy then asked how the U.S. could incentivize domestic AI investment while deterring non-state actor threats. Mr. Altman pointed to Project Stargate and emphasized clear regulation, talent, and supply chain access. Dr. Su advocated for more U.S. infrastructure, simpler export controls, and broader global adoption of U.S. tech.

  Sen. Sheehy asked if companies are choosing between the U.S. and China. Dr. Su confirmed that nations prefer U.S. tech but may seek alternatives if access is restricted.

• Ranking Member Cantwell asked if NIST should adopt AI standards. Mr. Altman said it could help but isn’t essential; the rest said “yes.”

  Ranking Member Cantwell then asked Mr. Smith and Mr. Intrator to elaborate on NIST’s role. Mr. Smith said NIST validates industry standards to support global leadership. Mr. Intrator noted standards enable common language and faster deployment.

  Ranking Member Cantwell warned about a Huawei-like scenario and asked about U.S. trade strategy. Mr. Intrator said demand will be met regardless, and U.S. leadership is preferable. Dr. Su supported controlled but broad access to allies to build a strong ecosystem.

• Sen. Moreno confirmed witness agreement on the urgent need to expand domestic energy generation. Dr. Su emphasized fabs and data centers require large amounts of energy. Sen. Moreno asserted that high-performance chips must be built in the U.S. to stay competitive. Dr. Su agreed and added the full semiconductor ecosystem is needed.

  Sen. Moreno criticized solar and wind focus. Mr. Smith said the AI industry will require diverse energy sources and a broad strategy.

  Sen. Moreno asked whether the internet era protected children. Mr. Altman said it failed to do so and committed to stronger AI safeguards.

  Sen. Moreno asked about stablecoins and AI. Mr. Intrator said both technologies could reshape digital payments and unlock societal benefits.

• Sen. Klobuchar asked if a risk-based approach to AI regulation is appropriate. Mr. Altman agreed, saying it helps avoid misuse while supporting innovation.

  Sen. Klobuchar asked if consumer AI education is necessary. Mr. Smith said yes, adding that industry should contribute.

  Sen. Klobuchar raised concerns about AI hallucinations. Mr. Altman said accuracy is improving and users are adapting as models evolve.

  Sen. Klobuchar asked about detecting deepfakes.
  Mr. Smith said Microsoft uses AI to label and identify AI-generated content.

  Sen. Klobuchar asked about compensating news outlets. Mr. Smith said collective bargaining and legislative action are needed to balance rights.

• Chair Cruz asked if the U.S. or China is leading in AI and what’s needed to maintain U.S. leadership. Mr. Altman said the U.S. leads narrowly and must sustain innovation and infrastructure. Dr. Su agreed but warned China is catching up via alternative strategies. Mr. Intrator said permitting and power are long-term threats. Mr. Smith emphasized global trust in U.S. tech is key to staying ahead.

  Chair Cruz asked about harm from an EU-style prior approval system. Mr. Altman said it would cripple innovation. Dr. Su and Mr. Smith agreed, with Smith warning of slowed deployment. Mr. Intrator said state-level fragmentation would hurt infrastructure buildout.

• Sen. Schatz asked if self-regulation is enough. Mr. Altman said limited government policy is needed, using aviation safety as an analogy.

  Sen. Schatz asked if labeling AI-generated content would help consumers. Mr. Smith supported transparency tools like C2PA to show content origins. Sen. Schatz raised concerns about AI exploiting IP without compensation. Mr. Smith supported fair compensation and clear copyright rules.

  Sen. Schatz asked how AI could improve public services. Mr. Altman said AI agents could simplify taxes and healthcare. Mr. Smith said routine tasks like license renewals could be automated.

• Sen. Budd asked how permitting delays affect AI energy needs. Mr. Intrator said access to large-scale power is critical and constrained by slow permitting.

  Sen. Budd asked Microsoft how it secures power and what policy changes are needed. Mr. Smith said Microsoft invests in new grid capacity and cited the Army Corps wetlands permit as a major bottleneck.

  Sen. Budd asked how the U.S. can lead in open-source AI. Mr. Altman said both open and closed leadership are key and previewed a new OpenAI model for summer.

  Sen. Budd asked what policy could best support the ecosystem.
  Mr. Altman said affordable, abundant energy is the most important factor.

• Sen. Kim asked Mr. Smith how Congress can boost international AI adoption. Mr. Smith urged updated export controls and fewer adoption barriers.

  Sen. Kim asked if global use of the U.S. AI stack is the best metric. Mr. Altman said yes, because it builds trust and influence across the full stack. Sen. Kim asked if foreign infrastructure use boosts U.S. application adoption. Mr. Altman agreed, saying it reduces risk of insecure alternative stacks. Sen. Kim asked if the U.S. is doing enough in AI applications. Mr. Altman said U.S. companies lead globally in application deployment.

  Sen. Kim asked Dr. Su to grade U.S. AI talent development. Dr. Su said the U.S. has strong talent but needs more and should expand high-skill immigration.

• Sen. Schmitt asked why Europe’s regulatory model has struggled. Mr. Altman said pre-approval delays frustrate users and stall deployment. Sen. Schmitt asked how vertical integration affects U.S. advantage. Mr. Altman said product adoption and feedback loops are key differentiators.

  Sen. Schmitt asked about importing European-style content control. Mr. Intrator warned that fragmented EU rules hinder investment and said OpenAI and Microsoft are better suited to handle content issues.

  Sen. Schmitt then asked about content freedom. Mr. Altman said censorship is un-American and AI should empower broad user freedom, not follow old social media frameworks.

• Sen. Hickenlooper asked about performance and safety reviews for Copilot. Mr. Smith said Microsoft uses joint and internal review boards, engineering tests, and red teams.

  Sen. Hickenlooper asked about external standards-based safety reviews. Mr. Altman said external red teaming is already critical to OpenAI’s process.

  Sen. Hickenlooper asked about energy-efficient chip development. Dr. Su said AMD is innovating in chip cooling and power efficiency, using AI to accelerate cycles and scale designs.

• Sen. Curtis asked why states like Utah attract AI projects. Mr. Altman cited power, fast permitting, workforce, and cooperative state leadership. Sen. Curtis asked how to protect energy consumers as demand grows. Mr. Altman said the key is increasing power generation to reduce costs overall.

• Sen. Curtis asked what role tech firms should play in power infrastructure. Mr. Smith said Microsoft partners with utilities and even self-funds upgrades to avoid rate hikes for local communities.

  Sen. Curtis asked how ChatGPT supports small business growth. Mr. Altman shared an example of an Uber driver using ChatGPT to run a laundromat, showing real-world impact on entrepreneurs.

• Sen. Duckworth warned against proposed cuts to national labs and asked the panel to speak on the importance of public-private partnerships with labs like Fermilab and Argonne. Mr. Altman praised the national labs as vital partners in advancing scientific discovery, stating that OpenAI has given model weights to a lab in its only such partnership, and predicted breakthroughs at unprecedented scale and speed.

  Sen. Duckworth invited other panelists to weigh in on national lab partnerships. Mr. Smith emphasized the labs’ foundational role in American innovation, describing them and U.S. universities as the backbone of the country’s scientific leadership since World War II. Dr. Su added that AMD has partnered with national labs for over a decade, praising them as forward-looking institutions well-suited for public-private investment. Mr. Intrator said national labs are integral to AI progress and suggested a tour of AI facilities would reveal just how much foundational science originates in those institutions.

• Sen. Young emphasized that the AI race was not only about jobs and innovation, but about national security and values, and asked whether failing to set U.S. standards would lead other nations to adopt theirs without American input. Mr. Smith agreed, citing the example of privacy law, and said that without U.S. action, other countries would lead and U.S. companies would likely be forced to follow foreign regulations. Sen. Young then asked whether having a baseline U.S. framework would help shape other countries’ standards. Mr. Smith responded that it would, noting the importance of timing and flexibility to harmonize standards over time for global interoperability.

• Sen. Blunt Rochester reflected on public fears about AI and asked why OpenAI transitioned to a public benefit corporation and what influenced that timing. Mr. Altman clarified that OpenAI always intended to remain a nonprofit with a PBC arm for capital raising, and said the change was long-planned but took time due to legal and stakeholder discussions.

  Sen. Blunt Rochester asked what supply chain policies are needed to support semiconductor competitiveness and AI progress. Dr. Su stated that the full chip production lifecycle—from advanced wafers to packaging—must have a strong U.S. footprint, alongside global partnerships for resilience.

  Sen. Blunt Rochester then asked how the interdependence within the AI stack could create risks or opportunities. Mr. Smith said it creates more opportunities than vulnerabilities, enabling companies to focus on their strengths and emphasizing the need for more innovation especially at the application layer.

• Sen. Moran asked how to balance privacy protections with AI utility, particularly as users share more personal information with AI systems. Mr. Altman warned that AI systems would become deeply personal and said new privacy guarantees would be essential as AI becomes integrated into people’s lives.

  Sen. Moran shifted to cybersecurity and asked how Congress should allocate federal resources given AI’s dual role as both a cyber threat and a defensive tool. Mr. Smith said AI is already helping block attacks in real time, such as in Ukraine, and recommended strong funding for federal agencies and cybersecurity leaders to stay ahead of threats.

  Sen. Moran raised concerns about rural broadband access and asked how AI systems could be designed to serve low-connectivity regions. Mr. Altman replied that cloud-based processing with minimal data transfer makes it possible to deliver high-quality AI services even in areas with poor connectivity, although broader broadband deployment remains important.

• Sen. Luján asked about the importance of federal investment in foundational research and standards bodies. Mr. Altman said standards could accelerate innovation but cautioned they should emerge from industry first and then be adopted officially. Dr. Su emphasized the value of blue-sky research supported by federal investments, and argued that public dollars should be used to leverage additional private sector contributions.

  Sen. Luján asked about local community partnerships when building data centers, specifically regarding energy and water impacts.
  Mr. Smith responded that Microsoft data centers used closed-loop liquid cooling systems with minimal water loss and maintained more than 90 water replenishment projects—including one in New Mexico—aimed at ensuring water-positive outcomes. Mr. Intrator said his company focused heavily on extracting more compute per megawatt and cited liquid cooling as a major step forward in energy efficiency.

  Sen. Luján then asked all witnesses to confirm whether fast, affordable internet access nationwide was essential for AI to reach its potential. All responded yes.

• Sen. Lummis asked what a state-by-state AI regulatory patchwork would mean for competitiveness. Mr. Altman warned that a fragmented regulatory environment would be burdensome and could significantly impede progress, and called for a single light-touch federal framework.

  Sen. Lummis asked the panel to elaborate on permitting challenges.
  Mr. Intrator described the permitting process as “excruciating,” noting that delays severely hindered large-scale AI infrastructure projects, especially energy-related ones. Mr. Smith agreed on the need for consistency across administrations and gave the example of natural gas-powered generators in Cheyenne, Wyoming, as part of Microsoft’s adaptive strategy.

  Sen. Lummis asked if Microsoft was exploring small modular nuclear energy. Mr. Smith confirmed that they were, including in discussions with stakeholders in Wyoming.

• Sen. Rosen asked how the U.S. should respond to adversarial AI models, particularly China's DeepSeek. Mr. Smith said Microsoft blocked the DeepSeek app from internal and public use due to security concerns but modified the open-source model itself to remove harmful elements.

  Sen. Rosen then asked how OpenAI was addressing antisemitism in AI outputs and whether he would support a civil society-led benchmark to prevent hate content. Mr. Altman said OpenAI already collaborates with civil society groups and aimed to prevent antisemitic outcomes while protecting user freedoms.

  Sen. Rosen turned to data center security and hardware, asking about efforts to improve chip energy efficiency and cybersecurity. Dr. Su reported that AMD had improved chip efficiency thirtyfold in recent years and was prioritizing both continued power efficiency and hardware security in future chip generations.

• Sen. Sullivan emphasized that the AI race was fundamentally a national and economic security issue and sought agreement from witnesses on whether the U.S. currently led. Mr. Altman affirmed the U.S. was ahead but noted the lead was fragile and stressed the need for infrastructure, a domestic AI supply chain, streamlined permitting, light-touch regulation, and talent immigration.

  Sen. Sullivan asked whether America’s energy resources and venture capital giving advantage to China were relevant. Mr. Smith acknowledged U.S. energy potential but warned of China’s rapid development capacity. He urged the U.S. to retain global talent and ensure capital remains focused on domestic innovation.

• Sen. Markey raised concerns about the environmental costs of AI, particularly energy and water use in data centers. He asked whether the government should conduct ongoing studies on AI’s environmental impacts. Mr. Smith and Mr. Altman agreed, though Mr. Altman cautioned that the environmental conversation should be balanced with AI’s potential benefits. Sen. Markey also focused on algorithmic bias and discrimination, citing its real-world impact on marginalized groups. Mr. Smith and Mr. Altman acknowledged that algorithms can be biased and confirmed their companies were actively working to guard against discriminatory outcomes.

• Sen. Peters emphasized workforce development, referencing legislation he sponsored to grow AI talent. He recalled a past conversation with Mr. Altman, who warned that 70% of jobs could be displaced by AI. Mr. Altman reiterated that AI’s potential for disruption was tied to its speed and advocated for iterative deployment of tools to help workers adapt. He noted that programming had already been transformed, with developers becoming more efficient, not obsolete.

  Sen. Peters then asked about the importance of hardware-level open standards. Dr. Su stated that open ecosystems foster innovation, reduce security risks, and enhance supply chain resilience. She affirmed AMD’s commitment to open models, contrasting them with Nvidia’s closed systems.

• Sen. Fetterman’s raised concerns about how increased energy demands from data centers might raise electricity costs for Pennsylvania families, referencing a report suggesting rates could increase by 20%. He expressed support for nuclear energy as part of the transition to cleaner sources and asked whether Microsoft would commit to avoiding increased costs for local ratepayers. Mr. Smith stated that Microsoft follows two key principles: (1) matching electricity usage with equivalent new supply on the grid and (2) ensuring that Microsoft’s presence does not increase local power rates. He described Microsoft’s practice of proposing self-funded grid upgrades to avoid burdening local communities.

  Sen. Fetterman asked about the singularity and public concerns over runaway AI development. Mr. Altman acknowledged the excitement and unknowns surrounding AI’s rapid progress. He emphasized the need for humility and caution, calling this moment a potential turning point in human history that will require societal adaptation.

• Sen. Klobuchar discussed bipartisan legislation on deepfakes and digital replicas, referencing real-world abuse of likenesses including deepfake videos of celebrities. She asked whether industry supports protections against AI-generated impersonations.

  Mr. Smith affirmed that deepfakes are a growing concern. He emphasized the importance of AI tools to detect manipulated media and supported voluntary industry standards alongside legal protections for digital likenesses.

  Mr. Altman supported the legislation and emphasized the need for multiple strategies—guardrails on distribution, public education, and legal accountability—while recognizing the challenge of stopping content generation outright.

• Chair Cruz about competition with China’s DeepSeek Mr. Altman downplayed DeepSeek’s significance but noted its consumer app momentarily surpassed ChatGPT. Dr. Su and Mr. Smith said DeepSeek highlighted innovation under constraints, while Mr. Intrator said it raised national awareness of China’s AI capabilities.

  Chair Cruz asked about the AI diffusion rule. All panelists supported rescinding the rule. Mr. Altman advocated for simplified, strategic export controls that allow trusted use of advanced chips abroad while safeguarding national security.

  Chair Cruz asked about federal preemption or a learning period for state-level AI regulation. All witnesses supported a single federal approach, with Mr. Altman and Mr. Smith advocating for light-touch national rules to prevent fragmented state regulation that could stifle innovation.

ADD TO THE NIMITZ NETWORK

Know someone else who would enjoy our updates? Feel free to forward them this email and have them subscribe here.