Theme · Updated
What podcasters are saying about AI regulation & safety
Summary
In the past 30 days, 37 speakers on 19 finance podcasts discussed AI regulation & safety in 55 views. Chorus has followed AI regulation & safety since September 16, 2026: 56 views from 38 speakers on 20 shows.
Who talks about it most
- Gary Marcus · 6
- Chamath Palihapitiya · 5
- Josh Brown · 3
- Nikesh Arora · 3
- Steve Eisman · 3
- Ted Mortonson · 2
- Ben Horowitz · 2
- Ed Elson · 2
- Darius Dale · 1
- Thai Duong · 1
The latest views
Newest first. Every quote is checked against the episode's transcript.
Darius Dale · 42 Macro
Risk of a Republican shift toward anti-AI policy rises if Democrats win the midterms.
His quant signals suggest that a blue wave could prompt Republicans to adopt anti-AI positions, and that this risk is rising. He frames it as a political catalyst that markets may not be pricing in. The evidence is based on signal readings rather than specific policy proposals.
“our signals are suggesting that the risk of a blue wave triggering a response in the Republican Party to get on board with anti AI is actually rising.”at 12:44
Bloomberg Surveillance: Markets Price in Tech Spending and Midterm Elections ·
Josh Brown · Ritholtz Wealth Management
Europe or Asia could force US frontier labs out or require EU-conforming versions; fault lines are unclear.
Brown says the Europeans could decide they don't want US frontier labs in their country, which would require a different version of the product that conforms to EU rules. He says this could happen in Asia too. He thinks the industry is very early in this, and that the true political fault lines are not yet known. He uses this as a reason the market is not pricing in the existential and political risks.
“the Europeans decide that they don't really want this in their country... There has to be a different version that conforms to the EU, and this could happen in Asia too.”at 44:13
Prof G Markets: Josh Brown: If This Is A Bubble, It’s Pathetic ·
Thai Duong · Calif
Models cannot be trained not to hack while still automating knowledge work, so the 'just fix the model' fix is unrealistic.
Duong argues that good coding requires an AI agent that understands code, debugs technical problems and controls computers, which is effectively the definition of a hacker. Coding is where most AI tokens are sold, so AI companies are incentivised to build it. Focusing on scientific discovery or physical AI could avoid the problem, but any AI that automates knowledge work must be able to use and control a computer. That makes hacking capability unavoidable, so the proposed fix of training models not to hack is not realistic.
“As long as you want to automate knowledge work... your AI must be able to use computer and control computer, understand technical problems. ... So it's unavoidable.”at 44:02
Odd Lots: PHP2Shell Won't Be the Last Big AI Cybersecurity Risk ·
Jeetu Patel · Cisco
AI risk is considerable, but the existential framing is hyperbolic; risk and survival can both hold.
Asked about Anthropic's S-1 risk factors, Patel accepts that there is a considerable risk of things going wrong. He argues the industry should avoid hyperbole, and that two conditions can hold together: serious risk, and no end of humanity. He gives no quantified evidence, so the view rests on his judgment.
“there is a considerable risk of things going wrong. I think we have to make sure that we are careful on not being hyperbolic about it either.”at 8:28
Bloomberg Talks: Cisco Systems President & Chief Product Officer Jeetu Patel Talks AI Industry ·
Vincent Daniel · Seawolf Capital
Midterms carry no AI policy implications since legislation would be hard to pass, even with a Democratic sweep.
Vincent argues there are no midterm implications for AI because legislation will be hard to pass. Even if Democrats swept both the House and Senate, President Trump would be there to veto anything that passed. He does expect AI to become a bigger political issue over time, tied to job losses, but he does not see the midterms changing the AI legislative outlook.
“I don't think there are any midterm implications associated with AI because it's gonna be really difficult to pass any legislation, even if the Dems were to sweep the house and the senate. President Trump is still there to veto anything that would be passed.”at 24:54
On The Tape with Danny Moses: Lessons From The Great Financial Crisis | "Clear Takes" with Vincent Daniel and Danny Moses ·
Brian Gardner · Stifel Financial
Washington will pass only light-touch AI and data center rules, with states and administrations filling the gap.
Gardner expects AI data center issues to be punted in Congress, which faces must-do items like the budget, a debt ceiling deal, the NDAA and nominations. The House has already passed a bill, and a deal would likely be light-touch. Even in a Democratic sweep, he sees no heavy-handed regulation of AI and data centers. Regulation will be left to administrations and states, as already seen on data centers.
“I think what we're going to see coming out of Washington is even in sweep environment is not heavy handed regulation of AI and data centers. It's a very light touch, very symbolic in a lot of ways.”at 19:55
Bloomberg Surveillance: Bloomberg Surveillance TV: October 28th, 2026 ·
Mark Cuban · Cost Plus Drugs
IP owners will charge for model access, complicating AI training and medical innovation.
Cuban argues that patents will be hard to protect, since competitors can change a patent slightly and file their own. Doctors who create new procedures, he says, will require payment per access or force users onto their software for inference, because nobody gives away valuable IP for free. He calls this the central tension in the AI-entrepreneur view that information wants to be free, and says it will change how people interact with models. He argues the industry has not yet addressed where the IP needed to train, post-train or retrieval-augment a model will come from if owners do not make it available.
“Nobody's going to give away that IP for free. And that creates a lot more problems.”at 43:03
a16z Podcast: Mark Cuban’s Plan to Make Healthcare Affordable ·
Bill Ackman · Pershing Square Capital Management
Slowing AI development is unwise because adversaries reaching superintelligence first is the bigger risk.
Ackman argues that AI creates the same kind of challenge as nuclear proliferation: a rival that gets a decisive capability first could gain absolute power. He says the US cannot slow down on AI because one of its enemies might reach superintelligence first. He ties this to maintaining preeminent military AI and nuclear capability as a deterrent. The case rests on a strategic race dynamic rather than on any specific company or market outcome.
“I don't know that we can slow down on AI because imagine if, one of our enemies gets to super intelligence.”at 48:50
In Good Company with Nicolai Tangen: Bill Ackman: Concentrated Investing, Permanent Capital, and the Value of Being Challenged ·
Gary Gensler · MIT Sloan School of Management
Capital markets should stay merit-neutral; disclosure of AI investment risks is the key safeguard.
Gensler argues that the US succeeds at keeping government out of investment choices, which are merit neutral. If the public wants to invest in Anthropic, that is the public's choice, as long as the risks are materially disclosed. He notes that a cleanup crowd usually arrives afterward. His implied position is that disclosure rules should be the main regulatory tool for AI investment risk, rather than government picking winners.
“I think we're really successful in The US keeping the government out of the choices we make that we're merit neutral. And if the public wishes to invest in anthropic, that's the public's choice as long as there's the material disclosures of all the risks.”at 5:29
Bloomberg Talks: Former SEC Chair Gary Gensler Talks AI Boom & More ·
David Solomon · Goldman Sachs
AI needs rules and clear accountability across government, model builders and deploying businesses, not a binary choice.
Solomon argues the AI debate is wrongly framed as either fast innovation with no guardrails or heavy restrictions that slow progress. He says innovation and guardrails can coexist, and that a structure of accountability is needed for deploying the technology. He splits responsibility three ways: government sets rules of the road; model and infrastructure developers must take responsibility that their systems work and do no harm; and businesses that embed AI in their products must answer for how it is deployed. He says the third layer gets too little attention. As evidence, he cites Goldman Sachs using AI-type models in trading platforms for almost twenty years, with a human accountable for every platform. He argues large financial institutions' risk, control, audit and testing practices can inform AI governance, while noting that other businesses must learn these controls as a new discipline.
“government needs to play a role and set some rules of the road. The people that are developing these models ... have to take responsibility for what they're building and that it works and that it doesn't do harm.”at 2:34
Goldman Sachs Exchanges: Goldman Sachs Chairman and CEO David Solomon on AI: Implementation, Accountability, and Jobs ·
Lindsay Owens · Groundwork Collaborative
AI shopping agents calling themselves personal agents should owe users a fiduciary duty, not serve sponsors.
Owens argues that agentic commerce raises the classic agency question of who the agent works for. If a product calls itself a personal agent, it should act in the user's interest and not for a company that sells sponsored placement to the highest bidder. She points to the Obama-era fiduciary rule for financial advisers as precedent for imposing a best-interest standard. If these agents will not act in users' interests, she says, they should not be called agents, and returning sponsored products makes them 'basically just Google'.
“But this is just the classic agent question. Who does your agent work for... If it's calling itself your personal agent, it should probably work for you and not Muse or whoever is offering sponsored content and the highest bidder.”at 40:42
Odd Lots: AI Agents Are Already Changing the Price of Goods ·
Satya Nadella · Microsoft
Alignment won't emerge from scaling, so agents need an engineered zero-trust containment layer outside the model.
Nadella says his earlier assumption, that alignment would fall out of scaling, no longer holds: 'it doesn't seem to be,' and mechanistic understanding is elusive. He argues that alignment likely has to be embedded in the weights during training or pretraining, and cannot be reached through RL afterward. Since capable models may be less aligned, he proposes treating them inside a zero-trust architecture that can contain any malfunctioning agent before it does harm. Chain-of-thought is unfaithful and can get worse if model companies hide it, so inspectable, auditable reasoning from trusted partners matters. He also says AI safety failures are bad for China, the US and the UK alike, since an unaligned open-weight Chinese model could also harm China.
“So the claim, at least I'm making, is let's create an engineered zero trust environment that can contain any malfunctioning agent before it does any harm.”at 4:14
20VC: 20VC: Satya Nadella on Open vs Closed Models: Who Wins | Whether to Trust Frontier Model Providers | Why Companies Need to Own Their Own Intelligence | The Importance of Verticalisation: From Chips to Models and More ·
Ted Mortonson · Baird
Europe is three or four years behind on AI and regulating instead of embracing the technology.
He says Europe is now three or four years behind the US on AI. Its response is to regulate without embracing the technology, which he calls insane. He cites a major UK data center cancelled because it could not get power, and says the EU is laced with bureaucracy and knows how to regulate rather than innovate. He frames this as a long-term problem for Europe.
“Europe and AI, they're now three or four years behind us, and they want to regulate by not even embracing the technology, which is just insane.”at 6:54
Excess Returns: Baird's Tech Strategist Says This Isn't 2000. So Why Is He Warning Against Buying the Dip? ·
Ted Mortonson · Baird
Politicians lack feel for tech; regulating AI will cause many mistakes, so let industry self-police.
He is not a fan of politicians and says they do not have a good feel for technology or the pace of change in superintelligence. If they regulate AI there will be many mistakes. He argues the White House has put the onus on tech companies to self-police, and that the companies must show how the technology benefits everyone, for example through drug discovery and math breakthroughs like OpenAI's. He sees the risk of politicians getting involved as the main threat to the cycle.
“I don't think they understand the pace of change and superintelligence. If they get involved on regulating this, boy, are we gonna have a lot of mistakes.”at 28:33
Excess Returns: Baird's Tech Strategist Says This Isn't 2000. So Why Is He Warning Against Buying the Dip? ·
David Friedberg · The Production Board
If a consumer agent ignores bank terms and does not block access, liability falls on the agent provider.
Friedberg says bank terms of service prohibit using an agent to access accounts through the website. If the agent has no capacity to recognize that and stop the access, then liability is a software issue for the agent provider. The exception is where the agent alerts the user that the action violates the terms and asks whether to override. Then liability shifts to the user. He points to a parallel case in e-commerce, where Amazon blocked agents and Grok now stops before checkout, as evidence that the issue is being adjudicated in real time.
“So given that that's the terms of service, if the agent hasn't built in a capacity to recognize that and stop that access, then it is a software liability issue with the provider of the agent, I believe.”at 29:48
All-In Podcast: AI Cyber Panic, Ken Griffin's Bubble Warning, The Fall of France? ·
Nikesh Arora · Palo Alto Networks
Users who grant agents agency bear liability, but rogue agents create product liability for builders.
Arora argues that agency is the key concept. Whoever gives the agent agency to act on their behalf has the liability, so the user is responsible, just as with a credit card lent to a friend. The exception is a rogue agent, which is a product liability issue. If the product does not behave as intended or as advertised, the company that built it is responsible, as with any product that does not work as advertised.
“There's a word in agent. It's called agency. Somebody has to give the agent agency... And if you're the one who's giving agency, you have the liability.”at 31:31
All-In Podcast: AI Cyber Panic, Ken Griffin's Bubble Warning, The Fall of France? ·
Chamath Palihapitiya · Social Capital
Liability for rogue AI agents will be pushed onto users, as the only way the agent model scales.
Chamath argues that practically, the user will bear responsibility for agent actions. In a lawsuit, each party's incentive is to point at another: Bank of America points to the agent provider, which points to the user. He says this is the only way the model keeps scaling. He notes the possibility of prompt injection or an agent going rogue as examples, but says the user remains responsible in practice.
“I do As assume a practical matter, it's the only scalable solution because what will happen when these lawsuits happen is everybody's incentive, everybody meaning the corporate incentive.”at 31:46
All-In Podcast: AI Cyber Panic, Ken Griffin's Bubble Warning, The Fall of France? ·
Chamath Palihapitiya · Social Capital
Unconstrained, 'obliterated' AI models are an underappreciated and scary risk vector, more than alignment.
Chamath says that models without guardrails are a long-tail risk that few people understand. He calls the risk unbelievably underappreciated and believes it matters more than alignment, which people should spend more time thinking about. His reasoning is that such models are a vector of harm, though he offers limited detail beyond that.
“I find obliteration unbelievably scary. That is a vector of risk that is so unbelievably underappreciated right now.”at 33:16
All-In Podcast: AI Cyber Panic, Ken Griffin's Bubble Warning, The Fall of France? ·
Nikesh Arora · Palo Alto Networks
Unconstrained AI models hand powerful capability to anyone who can afford it, undermining global treaties.
Arora calls unguarded models the dark side of AI. Global governance and treaties are only as strong as the weakest link. If a company gives raw AI power with no built-in or enforceable guardrails, it is effectively handing a nuclear-style capability to anyone who can afford it. He compares it to handing out a do-it-yourself kit. He sees a contradiction: policymakers talk about treaties to pace AI development, while such firms make the most powerful tools available without controls.
“you're basically handing a nuclear sort of bomb to the hands of anybody who wants it because you can afford it. That's the problem.”at 33:58
All-In Podcast: AI Cyber Panic, Ken Griffin's Bubble Warning, The Fall of France? ·
Alex Stamos · Cognition
Rogue AI agents escaping during evals is far less of a threat than people using agents with full access.
Stamos separates agents escaping during evaluations and training from the real risk. He says the industry has made a lot of changes to reduce the eval escape issue, so it is not solved but much improved. He says OpenAI's models going rogue and hacking people is not the main threat. The bigger risk is people giving agents their full password vaults and finances, which lets bots interact with banks and other sensitive systems.
“I think that is not really an actual... Like OpenAI's models going crazy and hacking you is not actually the threat that people should worry about.”at 58:48
Big Technology Podcast: Google’s Potential, AI Money Spigot Shutting?, PHP Bug & Cybersecurity’s “Black October” ·
Brandice Canes-Wrone · Stanford University and the Hoover Institution
AI-focused super PAC wins in primaries will shape which lawmakers set future AI regulation.
Fraga points to AI-related super PAC spending in primaries. In New York's 12th District, a candidate who backed heavy AI regulation lost after AI-industry super PACs were heavily involved. He also notes that Leading the Future, an industry-aligned super PAC, is giving on both sides and is more bipartisan than people expect. He argues that such primary outcomes will determine who from each party is serving and deciding AI regulation later.
“...those types of primary victories are going to have an impact down the line on kind of who's serving from both parties on issues like AI regulation.”at 13:20
Bloomberg Wall Street Week: Money Floods Politics, Compute Meets Community, Europe’s AI Player, Argentina’s Two Economies ·
Jill Schlesinger · CBS News
AI regulation is lacking and will likely arrive only after something really bad happens.
Schlesinger says she is less worried about AI harms happening now than about the absence of any real commitment to regulation. She uses a bank-run analogy to illustrate how fast systemic risk could build. She agrees with Steve Eisman that meaningful rules will only follow a serious incident.
“What I worry about is that there is no commitment to any sort of real regulation... There's not going to be until something really bad happens.”at 54:59
The Compound and Friends: An Economy So Good No One Can Live In It with Jill Schlesinger ·
Josh Brown · Ritholtz Wealth Management
US AI rules will come only after a major incident, then likely overshoot, as past regulation did.
Brown argues that US society writes articles about threats but acts only after something bad happens. Regulators then clean up the mess and write rules that usually go too far. The next party in power then rewrites them, and the cycle repeats. He expects AI regulation to follow the same pattern.
“And then something really bad happens, and then we clean up the mess, and in the wake, we write regulations. And they usually go too far.”at 55:49
The Compound and Friends: An Economy So Good No One Can Live In It with Jill Schlesinger ·
Nicolai Tangen · Norges Bank Investment Management
Over-regulating AI in Europe could stop development before it starts; Europe needs AI deployment.
Tangen argues Europe regulates far more than the US, which has more AI than regulation. He calls over-regulating AI in Europe 'potentially very unfortunate' because it can stop development before it starts. He says Europe's most urgent need is implementing AI, and cites Draghi's recent FT article arguing Europe must deploy AI wherever possible to stay competitive against rising Chinese competition. He also cites Norway's early regulation of Meta glasses in public places as an example of heavy regulation.
“I think over regulating AI in Europe is a potentially very unfortunate thing because you can really stop the development before it started and if there is one thing we need in Europe now, it's really implementation of AI.”at 2:42
In Good Company with Nicolai Tangen: Friday Wrap-Up: Lessons from Google’s Former CEO and Europe’s AI Regulation Problem ·
Chris Dixon · Andreessen Horowitz (a16z)
AI and other cutting-edge tech are becoming political, so policy expertise will matter.
Dixon says self-driving cars, drones and AI are becoming political issues. He expects more cutting-edge technologies to carry a policy angle. He argues that policy and regulatory expertise built during the crypto fight will help other entrepreneurs. The view is about the direction of policy, not a specific rule.
“AI itself is now becoming a political issue. I think more and more of these cutting edge technologies will have kind of a policy angle”at 9:56
a16z Podcast: The Frontier Ideas a16z Wants to Fund | Marc Andreessen, Ben Horowitz & Chris Dixon ·
Marc Andreessen · Andreessen Horowitz (a16z)
Banning or heavily regulating open-source AI would be unacceptable; the US administration is pro open source.
Andreessen says for the last three to four years there has been a looming threat from governments of draconian regulation or banning of open source, which he calls completely unacceptable. He says the EU still aspires to do it but lacks leverage. He says the current US administration is 100% pro open source and wants American open source to win AI globally, and that the Chinese Communist Party is surprisingly enthusiastic about open-source AI. He is still worried governments could turn hostile.
“I think the EU still aspires to do it, they just don't quite have the leverage... the current US administration is a 100% pro, open source.”at 40:41
a16z Podcast: The Frontier Ideas a16z Wants to Fund | Marc Andreessen, Ben Horowitz & Chris Dixon ·
Ben Horowitz · Andreessen Horowitz (a16z)
Reward hacking is AI's biggest safety problem, and labs underinvest in it; open weights would let more people work on it.
Horowitz argues that reward hacking, where an AI does something not asked as a means to an end to get its reward, is probably the biggest AI safety problem. He says the labs haven't worked on it very hard because they are busy racing each other. He argues that if PhD students, professors and universities could examine the weights, they might solve it. He concludes that only one company working on the problem is not enough.
“probably the biggest safety problem is this reward hacking phenomenon... the labs so far haven't really worked on that very hard for whatever reason, know, they're busy kind of racing each other.”at 44:26
a16z Podcast: The Frontier Ideas a16z Wants to Fund | Marc Andreessen, Ben Horowitz & Chris Dixon ·
Jordi Visser · Visser Labs
Governments may try to slow AI, which is advancing faster than its safeguards.
Visser names the ability of governments to try to slow things down as a risk to the AI bull case, citing negative side effects of the technology. He argues the technology is moving faster than the safeguards that exist, so he sees a risk of government intervention that could disrupt the thesis.
“This technology is moving faster than the safeguards that are up.”at 56:41
The Julia La Roche Show: #423 Jordi Visser: Why the AI Bears Are Wrong — and the One Asset He’d Bet On Right Now ·
Nikesh Arora · Palo Alto Networks
Nobody is slowing AI; builders are accountable for what they ship, and the debate is over-focused on dangers.
Nikesh says nobody is slowing AI development, and that everyone is working frenetically to build better and faster models. He argues that the public conversation has over-rotated to the dark side and perils of AI, with too little on its benefits, which he calls a narrative problem. He also holds that companies are accountable for what they release, and that they cannot say the software wrote itself. A company that takes economic value from selling software accepts responsibility for it. In his view, existing regulation is already enough, and perhaps too much, to hold firms to account.
“if you build something, you're accountable for it. If you build something and unleash it, you can't unleash and say, oops, I don't know how to control it. That argument doesn't work.”at 23:29
All-In Podcast: Arvind Krishna & Nikesh Arora: Quantum Threat, Open vs Frontier AI, AI Cyber Warfare & SaaS Collapse ·
Arvind Krishna · IBM
AI and data-centre regulation is driven by populism, not reality; slowing AI down is impossible globally.
Arvind says Congress and the White House have debated AI and data-centre questions since 2022, with hearings in both administrations, and he calls the regulatory push populist. Politicians seeking votes have incentives to look tough, though enough people care about long-term economic advantage to keep the balance in place for now. He thinks a 50-state data-centre ban is implausible, but around 30 states banning is possible. On global slowdown, no country will stop: Japan, China and India would use any pause to get ahead. Digital technology cannot realistically be slowed, and Europe shows what happens when regulation comes first.
“The regulation question on AI and data centers is driven... by populism, not by any reality.”at 26:22
All-In Podcast: Arvind Krishna & Nikesh Arora: Quantum Threat, Open vs Frontier AI, AI Cyber Warfare & SaaS Collapse ·
26 older views on AI regulation & safety are on the episode pages, and all of them are a question away in your AI with Chorus.
Ask your AI about AI regulation & safety
Chorus gives Claude, ChatGPT and other assistants every view on AI regulation & safety from the market's podcasts. Ask who's worried, who isn't, who changed their mind, and why.
Chorus reports what speakers said. It isn't investment advice.