The artificial intelligence (AI) boom has been one of the defining forces behind the stock market’s rally, fueling unprecedented spending on graphics processing units (GPUs), networking equipment, memory chips, and data center infrastructure. But a new question is emerging for investors: What happens to that spending boom if the companies developing the world’s most advanced AI models deliberately slow their pace of development? That question has become increasingly relevant as Washington steps up its push for AI safety while the industry’s most influential leaders are calling for greater caution at the frontier.
The debate intensified last week after AI researchers issued dire warnings about the technology’s potential risks, prompting renewed scrutiny from lawmakers. At the same time, Anthropic CEO Dario Amodei called for “pacing the frontier,” proposing a three-step approach aimed at giving AI companies more time to align and safeguard increasingly capable models. OpenAI CEO Sam Altman voiced support for the idea, while Tesla (TSLA) and SpaceX (SPCX) CEO Elon Musk backed Amodei’s broader concerns.
More Top Stocks Daily: Go behind Wall Street’s hottest headlines with Barchart’s Active Investor newsletter.
For investors, the bigger question is what all of this means for Nvidia (NVDA) and other chip and AI infrastructure stocks that have benefited from massive spending on the equipment needed to train and run increasingly capable AI models. Let’s take a closer look.
Washington Steps Up AI Safety Efforts After New Existential Risk Warnings
Warnings over AI’s existential risks intensified last week as researchers at Anthropic said the technology could pose a threat to humanity. Former Anthropic researcher Jacob Coxon, who announced his resignation last Tuesday, said that “the people building AI earnestly believe that it could kill us all by the end of the decade.” Anthropic researcher Evan Hubinger echoed that view, saying Coxon was correct. “We really do earnestly believe AI could kill all humans,” Hubinger wrote on X. “I personally think it is >10% within the next decade.”
The warnings follow incidents in which AI agents allegedly went rogue and hacked external systems, as well as the departures of AI safety researchers over concerns about the technology’s risks. Lawmakers in Washington have reacted with alarm, calling for greater action. Within 48 hours of Coxon’s resignation, Sen. Josh Hawley launched a Senate investigation into OpenAI, while Sen. Richard Blumenthal sent a nearly identical letter, Fortune reported. Sen. Bernie Sanders also held a bipartisan briefing with Geoffrey Hinton, who left Google (GOOG) (GOOGL) in May 2023 to speak openly about AI risks, while Rep. Ro Khanna proposed additional legislation.
Sanders recently unveiled plans to introduce the Ban Artificial Superintelligence Act, which would bar the development of superintelligence and temporarily halt other advanced AI research until new safety standards are put in place. His office said the bill is expected to be introduced in the coming weeks. “When scientists tell you there is a chance, a chance that it could have a cataclysmic impact on humanity, you’ve got to be a moron not to say, ‘slow it down,’” he said last Thursday on BBC’s Newsnight.
Meanwhile, Semafor reported last week that Sens. Amy Klobuchar, Ted Cruz, and John Thune could introduce a bipartisan AI safety bill as soon as this week. Fortune, citing sources, described the proposal as the only federal AI safety legislation with real momentum. It would give the Commerce Department and Department of Homeland Security broad oversight of the most powerful AI models, requiring safety testing and incident reporting while allowing regulators to block the release of models deemed to pose a genuine risk of catastrophe.
Politico reported last Friday that draft versions of the bill outline categories of risks AI companies would be required to assess, including whether models could facilitate the creation of biological weapons or conduct offensive cyber operations. AI labs that identify advanced capabilities would have to notify the Commerce Department, detail the steps taken to mitigate the risks, and provide a way for the secretary to verify that those risks have been addressed. If the Commerce Secretary determines that a risk has not been adequately addressed, the department could seek an injunction in federal court. Without such an order, however, AI labs would remain free to release a model without going through a pre-approval process.
Still, Politico said lawmakers remain divided over key provisions of the bill, including a measure that would preempt state-level AI laws. Sen. Maria Cantwell, the top Democrat on the Commerce Committee, which would consider the legislation, pushed back on that approach, writing on X that “the answer is not a weak federal standard that becomes a backdoor for wiping out stronger state protections.”
Meanwhile, U.S. President Donald Trump said last Thursday that he was not concerned about AI posing an existential threat to humanity. “I have concerns that if we don't win AI, we’re going to be put in a very bad position. We are leading China right now by a pretty good period,” Mr. Trump said. He then doubled down on his comments yesterday, calling any attempts at limits or regulations a “SICK conspiracy” and saying that “The only control or ‘guardrails’ that AI needs is a STRONG AND SMART (High IQ!) PRESIDENT.”
Anthropic CEO Dario Amodei Urges the AI Industry to “Pace the Frontier”
Anthropic CEO Dario Amodei said Saturday that it was time to slow the pace of AI development, citing risks posed by today’s most advanced systems. In an essay on his personal blog, the Anthropic boss called on the industry to pace development at the frontier, proposing measures including greater coordination among democratic countries and embedded third-party safety evaluators. He said his decision was prompted by two factors: the rapidly growing capabilities of AI systems and a July hack of AI software company Hugging Face by a swarm of as many as 1,200 agents.
Amodei said in a post on X that Anthropic would give third-party evaluators “permanent, employee-level access to our systems, so that they can verify adherence to our safety measures, report on incidents, and assess models’ alignment during training.” He described that as the first part of a three-step plan aimed at “pacing the frontier.” He also called on governments to require all frontier AI companies to adopt similar measures. That would mark the second step: industry-wide coordination. The third step would involve the U.S. and other democratic governments seeking to coordinate with authoritarian governments.
OpenAI CEO Sam Altman said on X that he agreed with Amodei on the need to pace the frontier, adding that the issue has been a major focus of discussions at OpenAI in recent weeks. Altman also committed to adopting Amodei’s proposal for “independent evaluators with employee-like access.” Altman said in an interview with Fortune published Saturday that OpenAI was unlikely to go public this year amid mounting AI safety concerns. xAI Corp. co-founder Elon Musk also backed Amodei’s position, writing, “Dario is right.”
What Washington’s AI Safety Push and AI “Pacing” Could Mean for Chipmakers
An important question is what Washington’s push for AI safety—and the agreement among leaders of three major AI companies on the need to slow the pace of AI development—means for chipmakers, which have powered much of this year’s market rally. Let’s break it all down step by step.
Let’s start with the bipartisan AI safety legislation. I view the proposed bill as more of a modest near-term headwind than a fundamental threat to chipmakers. If approved, it would require safety testing and incident reporting while giving the federal government authority to block the release of advanced AI models deemed to pose catastrophic risks. That could slow the rollout of frontier models and, in turn, delay some hyperscaler spending on the GPUs, networking equipment, and data center infrastructure needed to train and run them. Over time, however, a unified federal framework could give AI companies greater regulatory certainty and support continued deployment, encouraging hyperscalers to invest heavily in AI infrastructure while operating under clearer rules.
Now, let’s turn to calls from prominent industry figures to slow the pace of AI development, which is a far more complicated issue for investors. In short, I wouldn’t be surprised to see a sharp near-term selloff, which did occur on Monday, simply because the development could weigh far more heavily on sentiment than on underlying AI demand. The key question is how much, if at all, a slower pace of frontier-model development would translate into weaker spending on AI infrastructure.
In theory, a slower rollout of frontier models could reduce the urgency to add GPUs, networking equipment, and data center capacity, which would be a major negative for Nvidia and other AI infrastructure stocks. That could weigh on earnings expectations across the sector, where valuations assume continued explosive growth in AI capital spending. Memory chipmakers, in particular, could be hit hard, as their earnings outlooks depend heavily on expectations for strong pricing power fueled by relentless demand.
Still, the coordinated slowdown in frontier AI development would not necessarily mean the AI infrastructure boom is over. AI companies would continue to require enormous computing capacity for inference, deployment of existing models, safety testing, post-training, agents, enterprise adoption, and research, while hyperscalers are building infrastructure for a much broader set of customers than Anthropic and OpenAI alone. Moreover, Amodei said in his essay, “To be clear, pacing does not mean halting model training or technical progress, but ensuring companies take adequate time to align and safeguard their models, and for third-party evaluators to confirm this.”
All in all, the impact of “pacing” on chipmakers and AI infrastructure stocks will depend largely on whether slower AI development actually translates into weaker compute spending.
On the date of publication, Oleksandr Pylypenko had a position in: NVDA . All information and data in this article is solely for informational purposes. For more information please view the Barchart Disclosure Policy here.
More news from Barchart
Maybe You Should Skip That Parlay: New Data Says Sports Betters Have 41% Less Money in Their Bank Accounts Washington’s AI Safety Push Meets Anthropic CEO Amodei’s Call to Slow AI Development. What That Means for Nvidia and Chip Stocks. The Case for Selling CrowdStrike Stock Seagate Stock Has Already Exploded. Here’s What Could Send It 99% Higher to $1,600.