Episode summary
The Unexpected Investment Case for AI Safety
Thoughts on the Market (Morgan Stanley)
· · 10 min · Hosts: Ariana Salvatore, Michael Zezus
Summary
The discussion argues that tighter AI safety scrutiny is more likely to raise compute and capex spending than to slow it. The host and a guest say labs responding to safety concerns will need more monitoring infrastructure, which should increase demand for compute even as model capabilities rise. They also see limited odds of a sweeping federal AI law in the near term because of politics, procedure, and the historical need for a high-salience trigger.
The guest says the administration is unlikely to pursue an overarching AI regulatory authority or sweeping open-weight model restrictions. The reasoning is that the U.S. wants to preserve its managed-stability relationship with China, encourage diffusion of open-weight models, and avoid rules that would be hard to enforce once model weights are published. The guest thinks the executive branch can still set safety standards and guardrails, but not at the level of proposals like the Kill Switch Act.
On data centers, the guest says backlash is still driven more by environmental and affordability concerns than by AI safety itself. The buildout may become more conditional and expensive in certain jurisdictions, but the team remains constructive on AI capex this year and next. Midterms matter for sentiment, but the guest says a major policy shift would still require a triggering event more than a particular election outcome.
Key takeaways
- The guest argues AI safety scrutiny is a tailwind to compute spend and broader AI capex, not a brake on it.
- The team does not expect an overarching AI regulatory authority or sweeping open-weight restrictions in the near term.
- Data center opposition is described as mostly local and idiosyncratic, with environmental and affordability concerns leading the backlash.
- The guest says midterms may affect sentiment, but a major federal response would likely require a high-salience event.
- The recommended exposures are inference compute bottlenecks, cybersecurity leaders, and AI adopters, while avoiding negative exposure to rising U.S.-China tech-transfer restrictions.
Notable quotes
“We actually think this is a a pretty big tailwind to compute spend and capex more broadly”
“as the labs integrate greater safety monitoring infrastructure, we think that spend is only going to accelerate”
“We don't expect an overarching AI regulatory authority in the near term”
“we think this is a highly idiosyncratic issue, meaning that this is something to pay attention to on a very granular level”
“we still think that you should own inference compute bottlenecks because of that excess demand relative to supply”
How the episode opens
Welcome to Thoughts on the Market. I'm Ariana Salvatore, head of US public policy research at Morgan Stanley. — And I'm Michael Zezus, deputy global head of research at Morgan Stanley.
Today we'll be talking about AI safety and regulation. It's Wednesday, September 23rd at 10:00 a.m. in New York.
We put out a note last week on AI Frontier capability gain and the associated safety risks. Those have been in focus in recent weeks and as a result we've gotten a number of questions about the path forward for government regulation. So today Mike and I are going to get into some of the newest developments, where we think things are headed and how the midterms could shape that path.
Yeah. And this is pretty important because the concern is that if AI safety scrutiny increases, it's going to slow everything down. You might have less capex, fewer model releases, and there's all sorts of downstream effects for the pace of US growth and investment strategy in equities and throughout the AI investment theme, but Ariana, you and the team landed in a bit of a different place and are arguing that a bigger focus on AI safety could end up being a tailwind to compute spend rather than a break on it.
Can you break that down for us?
Sure. So, the way we see this playing out is there are five potential states of the world. Some include industry self-p policing.
Some include the prospects for heavier government intervention across all of them, as you mentioned. We actually think this is a a pretty big tailwind to compute spend and capex more broadly. That's because as the labs integrate greater safety monitoring infrastructure, we think that spend is only going to accelerate, especially as LLM capabilities increase at a nonlinear rate.
From the episode’s captions: roughly the first two minutes, lightly cleaned. Keep watching on YouTube →
Topics
- ai safety regulation
- ai investment pace
- compute spend
- capex outlook
- government regulation
Companies and tickers mentioned
- MS
- Morgan Stanley
Watch the episode
Watch “The Unexpected Investment Case for AI Safety” on YouTube
Summary, takeaways and quote selection by Octagon. Quotes are short excerpts; the full conversation is in the episode.
More from Thoughts on the Market (Morgan Stanley)
Thoughts on the Market (Morgan Stanley), this episode and its artwork belong to Morgan Stanley. Octagon is not affiliated with the show; this page is an independent summary.
