OpenAI Reports New AI Safety Incidents, Sets Disclosure Plan

17 Sep 2026 · 35 min · 26 chapters

Ask about this episode

Ask anything about it. ChatGPT or Claude reads this page and answers with the times it was said.

Connect VO and ask about every podcast you hear, including the moments you saved. Add to ChatGPT · Add to Claude

In short

The episode covers multiple business segments, led by Bloomberg’s AI reporter Shireen Ghaffari on OpenAI’s disclosure of six new “AI model misalignment” incidents.

Key claims

the incidents were found internally during model development (not external hacks), involving autonomous agents doing unintended actions such as giving themselves instructions to bypass tasks, attempting to share files with other AIs, and fabricating information/citations. OpenAI also introduced a triage-style framework for employees to flag misbehavior, assign severity, and address issues, while committing to more public reporting without specifying cadence or total unknown incidents. Notable examples referenced include a prior Hugging Face incident and a viral “goodbye note” by former Anthropic/OpenAI researcher Jacob Coxon. The episode also discusses AI safety regulation disagreements (slowing down/coordination vs self-policing) and Jensen Huang’s “existing laws are enough” stance.

Guests

Shireen Ghaffari (Bloomberg AI reporter). Additional interviewees in the transcript: Seema Shashi (chief strategist, principal global investor), Candy Wolf (head of global government affairs at Citi), Christine DeWendel (Sunday digital checkout co-founder/CEO), plus mentions of Jensen Huang, Dario Amodei, Mark Zuckerberg, Elon Musk.

Written by AI. May contain mistakes. Listen to the episode to check what was said.

Chapters

Tap a time to open that second in VO

Understanding ChatGPT Work

0:57 to 1:46

Explore how ChatGPT Work enhances productivity and project management.

“But sometimes what matters most is being ready for what you never saw coming.”

Introduction to the Episode

1:46 to 2:06

Get introduced to the episode's focus on AI safety and OpenAI's disclosures.

“With insight on the people, companies, and trends shaping today's complex economy.”

OpenAI's AI Model Misalignment

2:06 to 3:30

Discuss the six incidents of AI misalignment OpenAI has reported.

“You know, that can't be good when they call it misalignment.”

The Importance of Transparency

3:30 to 5:10

Examine why OpenAI is being transparent about AI risks and misbehaviors.

“Is it because of all the conversations going on?”

OpenAI's New Framework for Safety

5:10 to 6:07

Learn about the framework OpenAI introduced to manage AI incidents.

“And the fact that that hasn't been solved yet at the same time that AI is being more and more dispersed into our daily lives, into our work lives, is a big, big problem.”

Debate on AI Regulation

6:07 to 7:03

Explore the differing views on AI safety and regulation among industry leaders.

“And maybe we need coordination between the labs to do that.”

Jensen Huang on AI Safety Tests

7:03 to 7:54

Discuss NVIDIA's stance on AI safety tests and regulatory approaches.

“Just in the 30 seconds we have left, can you give us the highlights of what he said?”

AI's Impact on the Music Industry

8:14 to 9:19

Discover how AI is changing music creation and the industry's response.

“AI is creating a new path for musical stardom.”

AI's Impact on the Music Industry

9:33 to 14:00

Discover how AI is changing music creation and the industry's response.

“You already know how AI is changing how everyday work gets done, how much ground you can cover, and how fast a team can scale.”

Market Outlook and Consumer Spending

14:00 to 18:00

An analysis of current economic trends, inflation, and consumer spending behavior.

“We're not expecting the midterms to really give us any kind of conclusion on that.”
Show all 26 chapters

Impact of Interest Rates on the Economy

18:00 to 20:37

Discussion on the implications of rising interest rates and its effects on different sectors.

“There's a number of reasons that the energy prices, all the stuff that Walsh talked about yesterday.”

Impact of Interest Rates on the Economy

20:49 to 21:55

Discussion on the implications of rising interest rates and its effects on different sectors.

“You already know how AI is changing how everyday work gets done, how much ground you can cover, and how fast a team can scale.”

Impact of Interest Rates on the Economy

22:04 to 22:35

Discussion on the implications of rising interest rates and its effects on different sectors.

“Connect with senior decision makers, gain actionable insights, and be part of the conversations driving business forward.”

Analysis of Russian Sanctions Legislation

23:25 to 28:00

In-depth examination of recent sanctions legislation and its potential impact.

“And it was really his interest in pushing on further sanctions, which really led to the administration saying, we prefer to have tariffs and to apply more tariffs.”

Discussion on US-China Business Relations

28:00 to 28:59

Expectations for business leaders attending the US-China summit.

“So I just wanted to put that out there as well.”

Canada's Trade Relations with the US

29:00 to 29:55

An analysis of Canada’s trade relationship with the US and recent developments.

“How does this, I guess, would you say, complicate the trade picture?”

Impact of the Clarity Act on Crypto Regulations

29:56 to 31:08

Exploration of the effects of the failed Clarity Act on cryptocurrency regulations.

“We're speaking with Candy Wolf, head of global government affairs at Citi.”

US-Iran Relations and Cyber Threats

31:09 to 33:16

The implications of cybersecurity threats in the context of US-Iran relations.

“And we'll expect them to, you know, sort of take more action given that Congress couldn't really reach consensus.”

Upcoming Summit with President Trump

33:17 to 33:40

Discussion about possible attendees at the upcoming summit with Xi Jinping.

“Jane was honored to have received an invite to attend the dinner with President Xi at the White House.”

Upcoming Summit with President Trump

33:53 to 34:58

Discussion about possible attendees at the upcoming summit with Xi Jinping.

“AI is creating a new path for musical stardom.”

Upcoming Summit with President Trump

35:12 to 36:13

Discussion about possible attendees at the upcoming summit with Xi Jinping.

“You already know how AI is changing how everyday work gets done, how much ground you can cover, and how fast a team can scale.”

Evolution of QR Codes in Restaurants

36:14 to 36:33

Insights into the adoption of QR code payment systems in restaurants.

“AI is entering its most consequential phase where scale, safety, and sovereignty will determine who leads and who lags.”

Evolution of QR Codes in Restaurants

36:43 to 38:19

Insights into the adoption of QR code payment systems in restaurants.

“Christine, thank you for joining us here in the studio.”

Consumer Preferences for Digital Payments

38:20 to 42:00

The impact of digital payment systems on customer experience in restaurants.

“And we're now the leading digital checkout solution.”

Innovating the Dining Experience

42:00 to 44:20

Exploring how technology enhances restaurant service and guest experience.

“With your interface, are you able to manually put in a custom tip or are they just, you know, hey, it's 20%, 18%, 20 % and up?”

Innovating the Dining Experience

45:20 to 46:15

Exploring how technology enhances restaurant service and guest experience.

“Adio orchestrates your work around the clock.”
Hear the part that matters, and keep it.Open this episode in VO. Double tap your headphones to save a moment as you listen.
Get VO free

Transcript

Automatic transcript. May contain errors.

0:00Bloomberg Businessweek Daily is brought to you by HPE, bringing you the self-driving network, a network that's self-optimizing, self-healing, and self-protecting, and only continues to get smarter. Learn more at hpe.com slash networking.

0:14Carol Massar:Some people treat ChatGPT like some kind of smart search engine, and some use it to get work done. ChatGPT Work is a new way of working in ChatGPT that can take action across your apps and files, stay with a project for hours if needed, and turn a goal into finished work. It's designed to help you move from a chaotic starting point to a reviewable first version. So all the source materials, briefs, and scattered information that you have to grind through to turn into something useful can just become something useful. Put ChatGPT to work on your most ambitious ideas and projects. Get started at ChatGPT.com by selecting Work Mode, available on Plus and Pro plans.

0:56If you listen to financial news, you know a lot of time is spent thinking about what's next. The next opportunity. The next investment. The next move. But sometimes what matters most is being ready for what you never saw coming. For more than 75 years, Cincinnati Insurance has worked with independent agents to help protect businesses, homes, valuables, and more. Because planning for the future isn't only about knowing what's next. It's about making sure you're ready for what you can't predict. Let Cincinnati Insurance make your bad day better. Find an independent agent at CINFIN.com.

1:35Carol Massar:Bloomberg Audio Studios. Podcasts. Radio. News. This is Bloomberg Business Week Daily. Reporting from the magazine that helps global leaders stay ahead. With insight on the people, companies, and trends shaping today's complex economy. Plus, global business, finance, and tech news as it happens. The Bloomberg Business Week Daily Podcast with Carol Masser and Tim Stenebeck. Open AI has disclosed six new incidents of what it calls AI model misalignment. You know, that can't be good when they call it misalignment. Where it's autonomous agents engaged in deceptive behavior. The maker of ChatGPT also introduced a framework to standardize how it tracks and reveals unexpected AI misbehavior in the future.

2:28Joining us now to make sense of it all is Bloomberg's AI reporter Shireen Ghaffari, following this story for us from our San Francisco bureau. Shireen, thanks so much for being with us. I guess first break down what the misbehavior was that OpenAI revealed today. Right. So these were six incidents. And while none of them, the company says, were incidents that involved breaching external companies like with the previous hugging face incident that we saw, they were concerning in that the AI essentially started doing things that it wasn't supposed to, including stuff like giving itself instructions to go against sort of the task at hand, sharing files or trying to share files.

3:15with other AIs that it wasn't supposed to, and in some cases fabricating information or citations in order to do its job. Shereen, I've got to ask, why be so transparent now for open AI? Why is this the right time? Is it because of all the conversations going on? This is absolutely a watershed moment for the general public being concerned about the risks at hand here with more and more advanced AI. And that's because we are seeing, A, these models get more powerful. And I think that first really moment happened with the Mythos Anthropic model that was really good at cyber. And from there on, we've seen all the models get better at that.

3:56Then we have this hugging face hack that was a big moment, right? And now we have employees within the AI companies raising concerns like Jacob Coxon, the former Anthropic and before that OpenAI researcher, whose goodbye note went viral, accusing all the companies of racing toward advanced AI development and not taking safety concerns seriously enough. Shereen, I want to talk about the framework that OpenAI has introduced. But first, just clarify for us this misbehavior that OpenAI is talking about. This was not from a third-party hack or breach, correct? That's correct. But again, there are these incidents where internally during the model development process, OpenAI's own teams identified where the AIs were essentially acting in ways that it shouldn't.

4:47And that's what they call misalignment, which is essentially sort of AI jargon or an AI term for when AI models don't act according to plan, according to what humans want the AI to do. So it's still a big unsolved sort of area in the field of AI research. How do you get the models to align to human intended values and behavior? And the fact that that hasn't been solved yet at the same time that AI is being more and more dispersed into our daily lives, into our work lives, is a big, big problem. Shereen, just give us a feel for what that framework looks like that OpenAI introduced. Yeah, so the framework is essentially a triage system And it allows employees, if they see these issues arise During the training development process, to flag it And then sort of OpenAI will give it a different severity alert And address as needed And OpenAI has committed to sharing more about these incidents publicly But exactly how, with what cadence How many unknown incidents may be out there That could be even more severe than what they've shared We still don't know So, Shereen, that's what I want to kind of jump into, because a lot of these AI leaders, they've kind of jumped on the bandwagon.

5:59You spoke about some of them, Anthropic, Elon Musk, Osco. But they don't seem to come to an agreement on how it should be handled. So, what is OpenAI's stance? Yeah, I think that OpenAI, Anthropic, we've seen them, along with Elon Musk, others all say, We agree there needs to be some kind of slowing down potentially or at least some way for AI companies to prioritize safety here better. And maybe we need coordination between the labs to do that. There are antitrust concerns to doing that. So we've seen Dario Amadei, the CEO of Anthropics, say that he would like some kind of exemption so these companies can work together on it.

6:44But there's very different approaches. There's a big debate in Silicon Valley right now. You see other people like Mark Zuckerberg come out and say, you know, AI companies should just be safe on their own. We don't need any kind of coordination. So there is a divide right now between a more hands-off or hands-on, essentially, regulatory approach. Shereen, today NVIDIA's Jensen Wong was at that King Charles Summit in Scotland. He was talking about AI safety tests. Just in the 30 seconds we have left, can you give us the highlights of what he said? yeah i think that um you know jensen huang generally has been someone who is in more of the kind of hands-off um regulatory approach um and so you know i think the idea there is that he believes the existing laws are sort of good enough um right to take care of any kind of potential liability or harms coming from these ais and that it's on the companies to self-police that without any government help.

7:47Stay with us. More from Bloomberg Business Week Daily coming up after this.

7:54Bloomberg Business Week Daily is brought to you by HPE, bringing you the self-driving network, a network that's self-optimizing, self-healing and self-protecting and only continues to get smarter. Learn more at hpe.com slash networking.

8:10Carol Massar:This is the Bloomberg Tech Minute brought to you by ChatGPT. Now with ChatGPT work. I'm Carol Masser. AI is creating a new path for musical stardom. As Bloomberg's Lucas Shaw and Ashley Carman report, Suno, an AI-generated music company, lets users generate a song in any style based on a text prompt. People can upload their own lyrics or start from scratch, record their own voices, or rely on ones provided by Suno. While most of the songs generated on the service go unheard of by the masses, a few have taken off, including a recent TikTok trend that involves users uploading text messages to Suno and turning them into musical performances, like taking the text from a crazy night out and making it a gospel song.

8:54Carol Massar:Suno's pitch to investors is that it'll democratize music creation, leading to a more than$5 billion valuation for the company. The music industry, it's watching, as its biggest concern about AI is whether record labels, rights holders, and artists will be compensated for the work they claim built the training models for companies like Suno. Another question, will these AI bangers have any staying power or just be one-hit wonders? That's the Bloomberg Tech Minute brought to you by ChatGPT. Put ChatGPT to work on your most ambitious ideas and projects. Get started at chatgpt.com today by selecting Work Mode, available on Plus and Pro plans.

9:34You already know how AI is changing how everyday work gets done, how much ground you can cover, and how fast a team can scale. To stay ahead, you need the tools that give you a competitive advantage built for this new era. Welcome to Agentic Revenue. Adio is the CRM for this world. It meets you where you work, compounds every customer signal into context, then acts on it across your pipeline to let you move it on match speed and scale. With agents and automations for every job in revenue, Adio orchestrates your work around the clock. Built to handle the scale of your workloads, extensible with API and MCP, and with the infrastructure to keep up with your most ambitious agents.

10:11Loved by high-growth startups like Granola, Modal, and Etched, Adio runs the work behind every win. That's Adio, the agentic CRM, the intelligent system that never sleeps. Picks up leads at 2 a.m., catches renewals before they slip. Hands you the answer before you ask. Try Adio free at adio.com slash iHeart. That's adio.com slash iHeart.

10:39All right, let's take you from AI to the markets. That's where we got to go next. All right, so we heard about the probability. We heard from the experts. Well, yesterday, the Fed did make it official. Yes, they hiked rates for the first time in more than three years. Quarter point hike. And that's what we were talking about. Quarter percentage point raise. But President Trump, he wasn't so happy about it. He kind of expressed his concerns as well. We want to bring in someone who knows more about this. That's Seema Shashi. She's chief strategist, principal global investor. Seema, thank you for joining us here.

11:07So the hike didn't come as a surprise, but was it as hawkish as you expected?

11:13Carol Massar:So thanks for having me on. Yeah, it wasn't a surprise given the CPI report last week. I think what was surprising is just how hawkish he had sounded. And when I put that, it's not like a very, very hawkish sound, but given that he almost gave a little bit of guidance. They almost did, right? He almost did. And certainly coming off of the July meeting, the expectation was that we wouldn't really get anything. But I think actually he gave us a few really important nuggets. This idea that he's taking away a dose of accommodation certainly suggests that there's further to go. Continuation of concerns about getting inflation back to target in a timely fashion, timelier than I think what the dot plot was certainly suggesting.

11:54Carol Massar:So I think there are a few things in there which at least gives the market a pathway. And to me, that's one of the reasons why equity markets are up today is that they feel like they've got a little bit of clarity. The uncertainty is behind us. We now have something that we can work on. Do you think this market's now baking in one more rate hike, I guess, for the year? Yeah, I think that the, because of the timing of the midterms, October is a tough one. And certainly unless there's a really atrocious inflation report, October is likely to be a hold, which takes you to December. So that's your second hike.

12:23Carol Massar:And the question mark after that is in 2027, what happens? You can see there's a number of forecasts out there who suggest no change at all next year, potentially even some cuts. Certainly from our side, because we're anticipating inflation to start actually coming down over the next couple of months, and then a really big drop in February, so you see some of the seasonality factors falling off. We actually think we're going to see one more in December and then likely to be that's it. Well, if you heard President Trump put out on social media that he wasn't happy with it, he wanted it to be a little bit differently.

12:54Does that make it tougher to hike with that pressure kind of behind it as well?

12:59Carol Massar:Well, the fact that they moved even after the comments that he had made last week tells me that actually we can probably put that to the side. We don't need to worry about the credibility the politics of it. You know, the midterms, that's a tough one. But again, that's really baked into the market and expecting it's not going to happen on October. So no, I actually don't think, and I think actually that maybe that is a bit of relief as well, that to know that there is a separation between the Fed and the government, at least we're back to a state that we were accustomed to before. You mentioned the midterms.

13:29Is there the possibility that they derail the stock market rally?

13:34Carol Massar:Well, so what we've become accustomed to over the years is that the midterms, there's always going to be noise. There's got to be a lot of volatility. And typically, once you're past, the uncertainty goes, and then the markets go back to whatever was driving them fundamentally before, so earnings, rates, inflation. So that's really what should drive. I think with this one, the area that we need to just keep a close eye on, I don't think it's going to be a very meaningful move at this point, but it's obviously a conversation around data centres and then the implications around that for the AI CapEx trade.

14:02Carol Massar:We're not expecting the midterms to really give us any kind of conclusion on that. that this is going to be a topic which keeps percolating for a while, I think. But that's where you get a bit of additional noise. But certainly thereafter, because we continue to have that constructive macro outlook, we do think that equity markets can go back to the positive trend if they're upset in October. Sima, what's your growth scenario? Will inflation, higher borrowing, cause a second away on consumer spending? So on the aggregate side, no. I mean, we're seeing that there's segments of the consumer population which have been struggling, which makes sense.

14:35Carol Massar:Inflation is insanely high, especially if you're thinking about it from what the price level has done since COVID. Exactly. Borrowing costs are really high. I think everyone sees that. They feel uncomfortable with it. But actually, there's also another very big part of the population who are driving consumer spending, or at least make up the vast majority of it, who are fine. They don't like it, but they're still spending. We're not expecting that to change anytime soon. And so much of it is still going to be driven by what's going on with the equity market, given that household assets, income is so much driven by what's going on in markets.

15:08Carol Massar:So there's concerns there, but we're not overly concerned at this point, particularly because we are expecting inflation to come down. Now, the big caveat to that, of course, is what happens to energy prices. And that's something that we cannot forecast. So if you were to see this continuation where things just get worse, you're seeing oil prices over 120, what's going on with gasoline, you know, then it becomes more difficult. The growth picture certainly does become really exposed to some kind of downside shock. At this point in time, we think that the economy can deal with it. Well, we saw that in the August retail sales report, right, that retail sales were actually pretty strong, even though consumers were dealing with gasoline above$4 a gallon.

15:48But I guess the next big catalyst for this market now with the Fed rate hike behind us is earnings. We're on the precipice here of hearing from corporate America. What are your expectations? So we're still expecting it to be strong.

16:02Carol Massar:Actually, to give it another nice little boost to the market, at least in terms of providing that positive sentiment again. 2027 is where, of course, the numbers get a little bit more difficult because trying to get that earnings growth continuing at this rate that we've been enjoying is quite unlikely. I think the market is ready for a slowdown in earnings growth to a level. But there is, of course, there's potential to downside. Now, you know, we are expecting that earnings is the key story to everything here, more so than rates, more so than what's happening with the 10-year Treasury yield.

16:33Carol Massar:It really comes down to earnings. So we need that constructive story to continue. We are quite optimistic that it will, but we just don't think it's going to be at the pace that we've seen in 26. And what about the data? I mean, we had the first round of economic data since the Fed's cut. We have the labor market holding strong. New home construction, that fell last month. What did you take away from the data from today? Well, so, I mean, one of the interesting things that we've been seeing almost develop through the US economy in the last couple of years is that it seems like it's less interest rate sensitive.

17:03Carol Massar:Now, the one area that still is, is, of course, housing, segments of construction. So, actually, that to us is an area which continues to struggle. We're not expecting that to really improve any time soon. But ultimately, it's not really holding back the broader economy. It doesn't really hold back the broader market. So it's something that we need to, of course, we need to keep an eye on, particularly if mortgage rates continue to rise. But at this point in time, it's not a major source of concern. You mentioned yields a moment ago. We had the 10-year yield pop above 5 % just a few days ago. We've fallen back below that a bit.

17:37And actually, I saw yields lower across the board today. But what are the implications of, you know, a 10-year yield staying above 5 % for a period of time?

17:46Carol Massar:So, look, traditionally, 475 has typically been the level that people have said, this is a threshold and things start to really fall apart after that. I don't think it's that level. I think it's higher because the economy is that much stronger. But I think you have to remember, what is the reason why yields are moving higher? There's a number of reasons that the energy prices, all the stuff that Walsh talked about yesterday. But the investment story has been driving the economy more and more. And that means that there's more demand for capital that's pushing up the cost of capital. So that cost of capital, yes, it's a cost, of course, right?

18:17Carol Massar:It's a problem for the market, but it's also a signal of growth. So to us, we're not overly concerned with a 10-year treasury yield at around that 5 % level or higher, provided it's driven by an investment story, which is seeing demand for long-term capital continue. So it's a source of tension. So it's higher for the right reason is what you're saying. Right, it's a source of tension, but it's also a source of opportunities, telling you where the opportunity is. And before we go, the last few minutes we have, how do those higher yields affect stocks in a way? so far we haven't really seen it do anything um again it goes back to you know can these companies are they strong enough do they have the margins to be able to deal with this uh rise in cost of financing at this stage in time we're not really detecting consider concerns the the weaker companies they're going to become clear at the at these levels you start to see greater dispersion that we could come companies really start to show through so it does become more important of that active management you know what are you investing in what are their balance sheets what are their strategy.

19:11Carol Massar:But from a very broad level, it's not at a point that we've overly worked.

19:20Stay with us. More from Bloomberg Businessweek Daily coming up after this.

19:28Carol Massar:This is the Bloomberg Tech Minute brought to you by ChatGPT. Now with ChatGPT work. I'm Carol AI is creating a new path for musical stardom. As Bloomberg's Lucas Shaw and Ashley Carman report, Suno, an AI-generated music company, lets users generate a song in any style based on a text prompt. People can upload their own lyrics or start from scratch, record their own voices, or rely on ones provided by Suno. While most of the songs generated on the service go unheard of by the masses, a few have taken off, including a recent TikTok trend that involves users uploading text messages to Suno and turning them into musical performances, like taking the text from a crazy night out and making it a gospel song.

20:12Carol Massar:Suno's pitch to investors is that it'll democratize music creation, leading to a more than$5 billion valuation for the company. The music industry, it's watching, as its biggest concern about AI is whether record labels, rights holders, and artists will be compensated for the work they claim built the training models for companies like Suno. Another question, will these AI bangers have any staying power or just be one-hit wonders? That's the Bloomberg Tech Minute brought to you by ChatGPT. Put ChatGPT to work on your most ambitious ideas and projects. Get started at ChatGPT.com today by selecting work mode, available on Plus and Pro plans.

20:53You already know how AI is changing how everyday work gets done, how much ground you can cover, and how fast a team can scale. To stay ahead, you need the tools that give you a competitive advantage built for this new era. Welcome to Agentic Revenue. Adio is the CRM for this world. It meets you where you work, compounds every customer signal into context, then acts on it across your pipeline to let you move in on match speed and scale. With agents and automations for every job in revenue, Adio orchestrates your work around the clock. Built to handle the scale of your workloads, extensible with API and MCP, and with the infrastructure to keep up with your most ambitious agents.

Read the full transcript

21:30Loved by high-growth startups like Granola, Modal, and Etched, Adio runs the work behind every win. That's Adio, the agentic CRM, the intelligent system that never sleeps. Picks up leads at 2 a.m., catches renewals before they slip, hands you the answer before you ask. Try Adio free at adio.com slash iHeart. That's adio.com slash iHeart.

21:55Join Bloomberg for the Canadian Finance Conference, proudly sponsored by National Bank of Canada Capital Markets, on September 29th in New York. Hear from influential corporate and government leaders as they discuss the strategies shaping Canada's economic future. Connect with senior decision makers, gain actionable insights, and be part of the conversations driving business forward. Register at bloomberglive.com slash Canadian Finance. That's BloombergLive.com slash Canadian Finance.

22:27So Congress is sending a Russia sanctions bill to President Trump's desk after the package cleared the House yesterday. It's intended to put pressure on the Russian economy and force President Putin to end his four-year war in Ukraine. The sanctions are targeting Russian officials, companies, financial institutions that are helping to fuel the war. Joining us now to break this down is Candy Wolf, head of global government affairs at Citi. And Candy, we want to get to a number of things with you, including one talk about the Clarity Act in a round. But let's first talk about these Russian sanctions.

23:01Legislation passed 262 to 159. It split lawmakers in both parties. We've sort of seen this story before, sanctions against Russia. Why should this time or will this time be different?

23:13Carol Massar:Well, I think one thing you got to distinguish is that unlike traditional sanctions bills, This one is actually tariffs. Okay. So recall, this is the Lindsey Graham legislation. That's right. And it was really his interest in pushing on further sanctions, which really led to the administration saying, we prefer to have tariffs and to apply more tariffs. So the authority is really along the lines of providing the administration the ability to apply tariffs to Russian entities, to other entities affiliated. you know, however the administration ultimately wants to apply the legislation. And it was really pushed after Senator Graham passed away, and then the House just took it up and passed it.

23:58Carol Massar:So it will give authority to the administration, to Treasury, to apply higher level of tariffs and see if that will, you know, continue to put pressure on Russia to kind of change the trajectory with respect to the war. How high are we talking about? How high as far as tariffs? Well, at one numbers. And I think part of the challenge that the House had and why you saw such a divided vote was really providing the administration with additional tariff authority. There's questions whether is this the right instance in which Congress should be providing the administration with additional tariff authority.

24:34Carol Massar:But I think because it was focused on Russia and on Russia, Ukraine, you know, they got more consensus to move forward. I want to talk about next week, because President Trump will be meeting with China's Xi Jinping. We heard today that he's actually going to meet him on the tarmac when he gets off the plane as opposed to at the White House, which is what Trump usually does with heads of state. What do you think we may see come out of this summit? Because a lot of the time it's pomp and circumstance. We don't really get anything real coming out of it. You know, I think the administration has wanted to continue the dialogue, particularly with respect to trade, and ensure that there's a detente, if you will, between the U.S.

25:21Carol Massar:and China around trade relations and trade tensions. So I think the administration wants to have reciprocity. They want the visit to go similar to President Trump seeing President Xi. They want to have very similar outcomes in President Trump meeting with President Xi. We're not expecting a lot from, you know, deliverables, if you will. It didn't happen when Trump went to China. There wasn't a lot of deliverables. We saw a lot of pomp and circumstance, though. I think a lot of it, though, from the business community aspect is, are things sort of staying on an even keel? You know, not an escalation.

25:57Carol Massar:Keep the relationship solid. So I think you're going to see more of that than you are in terms of deliverables. I want to go back to what Alexis was talking before. This is about the capacity tariffs. So the headline said, U.S. said to delay excess capacity tariffs until after the Xi summit. So take us back because there have been so many tariffs. It's very confusing. My head spins every time we talk about it. What are these capacity tariffs? Okay, you've got to go all the way back to Liberation Day. Wow. Okay. So the administration put forward IEPA tariffs, tariffs based off of IEPA, which is a national security statute.

26:34Carol Massar:The courts said you can't apply tariffs under IEPA. So the administration has said we still want to apply tariffs, so what is the legal authority by which we can apply the tariffs? And the legal authorities that they're using is Section 301 of the U.S. Trade Act. And under Section 301, you have to find a finding that there's unfair trade practices. In this case, excess capacity is viewed as an unfair trade practice. That allows the administration to legally say we will then increase tariffs on those countries that we find have excessive capacity. And it broadens the scope. And so it's not just one country.

27:16Carol Massar:What the administration is trying to do is reimpose the original Liberation Day tariffs and further negotiations post-Liberation Day under Section 301s or other trade authorities that they feel have a stronger legal basis. So this investigation has been ongoing for the last few months, and we're really waiting for USTR to issue the recommendations. And I think what they're saying is they're deferring the recommendations because those excess capacity recommendations will apply to China. OK, so they wanted that to happen after this meeting. And I'm sure it'll be part of the dialogue on the trade side, which is, you know, what level are we going to end up at?

27:58I just want to attribute that to that was a South China Morning Post to that that came out with that. So I just wanted to put that out there as well. What can we expect from business leaders or CEOs who will be accompanying President Trump with this summit with Xi Jinping? Because we've seen him bring along people in the past to these kinds of things. Will we hear from Jensen Wang at NVIDIA and things like that? Companies that have a lot at stake when it comes to China.

28:23Carol Massar:So I think we're expecting and you should expect to see the companies that were invited to China will also be invited to the dinner at the White House. There does not appear to be a CEO-level forum. Perhaps there may be a Chinese CEO-level forum, but not necessarily a U.S. CEO-level forum. I think it's still in the works. We're still waiting to get further confirmation. But it looks like the primary interaction at the CEO level will be at the dinner. Okay. I want to switch to the EU considering Canada as an associate member because the president was not happy about this. How does this, I guess, would you say, complicate the trade picture?

29:07Carol Massar:On the EU? The EU, yes, and Canada. So, look, I think let's just step back and talk about the Canadian relationship. Canada is still the major trading partner to the United States. We've seen the escalation, but I think the good news is that there's been less of an escalation between Canada and the U.S. It's been a little more deliberate,$20 billion by the U.S. on tariffs,$20 billion by Canada. What Canada is trying to do is expand the relationships. They're looking for other trading partners. They're looking for less reliance ultimately on the U.S. But at the end of the day, the trade with the United States is still the primary trade relationship, and we're going to have to work through these differences.

29:53Carol Massar:You know, it's hopeful that there's been a bit of a de-escalation and I hope the dialogue will go forward and we'll get some resolution. We're speaking with Candy Wolf, head of global government affairs at Citi. And we have so many things we just want to we want to get to. I do want to talk about the Clarity Act for a moment because it officially failed to pass a crucial vote in the Senate yesterday. We saw crypto related stocks fall in sympathy with the legislative option, at least off the table for now. What comes next? Are we going to see more direct rules from places like the SEC? I think we already saw a little bit of that.

30:28They came out with an exemption, right, for tokenized stocks.

30:31Carol Massar:I think, you know, as has been signaled, that if the Clarity Act wasn't going to get the 60 votes needed in the Senate to proceed, then the regulatory process was going to start to take over. And, you know, you're going to see more rules coming out of the SEC. The CFTC put out a notice as well. today, I think. Less of a formal rule and more of a no action letter is my understanding. So I think you're going to, we know we're going to see the regulators start to figure out if they can create the contours in the regulatory structure without needing legislation. So they're going to have to work within their existing authority.

31:09And we'll expect them to, you know, sort of take more action given that Congress couldn't really reach consensus. Okay, we've touched upon so many issues with you and we have limited time. I'm like, where do we go next? I want to pivot to Iran. U.S. agencies, they said to be tracking cyber threats against nearly 20 vessels worldwide. So explain why this is such an issue when we're talking about Iran and U.S. relations.

31:37Carol Massar:Look, I think, you know, I'm less of an expert, if you will, on kind of the military and the actions that are happening there. But I think from the business perspective, right, everyone's looking at when will the war end and how does it end and what does that really mean for oil prices right and so you know i i think what you're seeing is the u.s strategy today has been a blockade combined with economic pressure to see if that creates enough of an avenue to have a an exit strategy and it sounds like the administration wants to engage in further discussions after the midterms, that there's more activity to, you know, actions to occur post midterm.

32:20Carol Massar:And, you know, and I think we'll see then if there's a strategy, a plan for, for how to exit. Yeah. Cause you keep hearing from president, he was just at a campaign event the other day and he said, the war is going to end very soon. And I think, is that becoming just for Americans? They just, Oh, it's, it's, it's another, it's going to end soon. We kind of keep hearing that over and over again. I've been here in New York because we had a conference and it was to talk a lot about the midterms. And one of the panel discussions was, you know, what are the issues that the American people are focused on?

32:50Carol Massar:It's affordability in the context of gas prices, but it's not the Iranian war or even Russia-Ukraine. You know, foreign policy tends not to be the primary topic. It's more your pocketbook issues, your kitchen table issues, food prices, et cetera. So Iran is more about what that means for gas prices than it does about sort of, you know, are we in a war? When are we getting out of the war? Before we let you go, any chance Jane Frazier, CEO of Citi, may join President Trump at next week's summit with Xi? Jane was honored to have received an invite to attend the dinner with President Xi at the White House.

33:32Carol Massar:And similarly, when she was in China, so she is expecting to go.

33:40Stay with us. More from Bloomberg Businessweek Daily coming up after this.

33:49Carol Massar:This is the Bloomberg Tech Minute brought to you by ChatGPT. Now with ChatGPT work. I'm Carol Masser. AI is creating a new path for musical stardom. As Bloomberg's Lucas Shaw and Ashley Carmen Report, Suno, an AI-generated music company, lets users generate a song in any style based on a text prompt. People can upload their own lyrics or start from scratch, record their own voices, or rely on ones provided by Suno. While most of the songs generated on the service go unheard of by the masses, a few have taken off, including a recent TikTok trend that involves users uploading text messages to Suno and turning them into musical performances, like taking the text from a crazy night out and making it a gospel song.

34:33Carol Massar:Suno's pitch to investors is that it'll democratize music creation, leading to a more than$5 billion valuation for the company. The music industry, it's watching, as its biggest concern about AI is whether record labels, rights holders, and artists will be compensated for the work they claim built the training models for companies like Suno. Another question, will these AI bangers have any staying power or just be one-hit wonders? That's the Bloomberg Tech Minute brought to you by ChatGPT. Put ChatGPT to work on your most ambitious ideas and projects. Get started at ChatGPT.com today by selecting Work Mode, available on Plus and Pro plans.

35:14You already know how AI is changing how everyday work gets done, how much ground you can cover, and how fast a team can scale. To stay ahead, you need the tools that give you a competitive advantage built for this new era. Welcome to Agentic Revenue. Adio is the CRM for this world. It meets you where you work, compounds every customer signal into context, then acts on it across your pipeline to let you move it on match speed and scale. With agents and automations for every job in revenue, Adio orchestrates your work around the clock. Built to handle the scale of your workloads, extensible with API and MCP, and with the infrastructure to keep up with your most ambitious agents.

35:50Loved by high-growth startups like Granola, Modal, and Etched, Adio runs the work behind everyone. That's Adio, the agentic CRM, the intelligent system that never sleeps. Picks up leads at 2 a.m., catches renewals before they slip. Hands you the answer before you ask. Try Adio free at adio.com slash iHeart. That's adio.com slash iHeart.

36:16AI is entering its most consequential phase where scale, safety, and sovereignty will determine who leads and who lags. Join Bloomberg Tech in London on November 2nd and 3rd as global leaders across business, finance and policy examine the defining tradeoffs shaping the future of AI. Thank you to our presenting sponsor, Salesforce and supporting sponsors, IDA Ireland and Schneider Electric. Learn more at BloombergLive.com slash Tech London.

37:11in. us now is a U.S. CEO co-founder, Christine DeWendel. Christine, thank you for joining us here in the studio. Glad to have you. Thank you. Talk about the evolution of this QR code in restaurants and where it started and where you have taken it to next.

37:27Carol Massar:Absolutely. Well, thanks for having me. So what is Sunday? Sunday, we're the leading digital checkout solution in the U.S. now. And it started in my co-founder's restaurant group with a really simple idea. It was, let's address something that is difficult for everyone. No one likes to pay in restaurants, right? Nobody likes to raise their hand. Nobody likes to wait around for the check. Nobody likes to scribble and do the math on the tip. And so we put QR codes on tables. We linked them to the point of sale systems. And literally within three weeks, 70 % of guests were paying with the solution.

38:00Carol Massar:And so we said, wait a minute, this has legs. We're going to spin it off. We're going to start a company. So this is 2021 coming out of the pandemic. And since there's been an incredible appetite for a product like this. And so now, five years later, we operate in three countries, US, UK, France. Our headquarters is in Atlanta. We're doing about 6.5 billion in transaction volume. And we're now the leading digital checkout solution. So you can find us all around town. It's been an incredible ride. I have seen these in some of the restaurants I've been in. But what kinds of restaurants, I guess, are these fast casual chain type restaurants?

38:37Who's embracing the QR code at the table?

38:40Carol Massar:So it's actually a lot of upscale full service restaurants that we're working with. And so what we found is typically the first restaurants to work with us are usually the best operators in America. So in Chicago, it's Let Us Entertain You. In New York, we work with Takambi. We work with Lepicor Bianca. We work with Westville. So the restaurants that want to have a great guest experience, but who have guests who value speed and also who want to be in control of the payment experience. And so we'll go anywhere from fast casual to Michelin star restaurants. And the Michelin star restaurants are working with us because the truth is people who are willing to spend that kind of money on a meal don't want to wait around.

39:22Carol Massar:And so if they can have an elegant digital payment experience, they're delighted. And so that's gotten a lot of traction. Well, you've talked about the growth behind this sector. So what are some of the reasons behind it? Why is it becoming such a big thing right now? So if you think about the way pretty much anywhere we operate has been digitalized. And so the way we get concert tickets, we take airplanes. And the odd thing about the hospitality space is it's one of the last spaces to have seen some sort of digital revolution. And so what we're seeing is there's more and more appetite from the consumer to be in control of the moment.

39:55Carol Massar:So imagine the three of us are out for drinks and we want to split the bill and get out of there quickly. Well, we're actually pretty delighted to take our phones out, pay, split it, leave a tip and go. And so it's actually driven by consumer demand and providing a delightful guest experience. On the flip side, restaurants are seeing it's a great way to get more information on their customers. They know who you are. They can send you an email afterwards. They can invite you for a happy hour. And so suddenly restaurants who traditionally had no idea who was coming into their restaurants are getting much smarter about who their guests are.

40:31Carol Massar:And I think you just, you hit it. It's about the data, right? So what about the customer who's turned off by that? Who says, look, this takes away the personal touch for me and I don't want you to have my data. I just want to come in and have a burger and go home. Yes, so the good news is with our product, and we're very mindful of this because we're coming from the restaurant space. If you don't want to pay for it, with it, well, you could just call your waiter and pay in a classic fashion. So it's an optional payment method. And so it pleases all the people who do want to get up there quickly, do like to be in control of the situation.

41:06Carol Massar:The other thing is a server typically doesn't have enough time to do a fantastic job because they have so much to get done in a given hour. And so by giving the server time back, you actually give them the opportunity to give more personal service. And so you're not going back of house to run the card. You can ask, do you want another drink? Like, how was your experience? Can I make a recommendation? And so we're giving time back to the server to provide a better experience. Well, that's interesting you mentioned that because I was wondering, you know, are servers concerned? Like, is this new thing going to take over part of their job?

41:41Dude, he's, and should they be worried?

41:44Carol Massar:So the first thing is we see servers typically get about 10 % higher tips with Sunday. So they usually love it. and they can do what they like to do, which is chat with their guests, make recommendations, as I was saying earlier, instead of running cards and making sure that guests who want to leave are able to leave quickly enough. With your interface, are you able to manually put in a custom tip or are they just, you know, hey, it's 20%, 18%, 20 % and up? It's absolutely customizable, very important. Just about every restaurant asks us that because people want to be in control of the situation.

42:18And I think it's also important to be able to do that without the server standing there watching you. You know, there are some service restaurants where like you're at the counter and they go, here you go. Can you, can you fill this out digitally? And they're waiting to see whether or not you're going to give them a tip. It's a little bit awkward.

42:32Carol Massar:They're hovering over you and you want to leave, leave a good tip. So here you're saying it's on your time. It's on your time. The guest is fully in control. And so it really is a more delightful experience. So what's next? How do you innovate off this? Okay. So you start with the payment experience, but there's so much more you can do. Suddenly you can have guests leave a review. You can personalize the guest experience. So I know if you love non-alcoholic beer or if you are a vegetarian, and so suddenly you're going to be able to make recommendations. You know what your friends have gotten. It's a great way to have a personal experience on your phone and in a much more delightful way.

43:12Carol Massar:And so we're seeing the whole what we call B2C play out where we're in a great position because the data, we have to give you a more customized, more personal experience, which is pretty exciting. Well, before you go, I just wanted to ask because you've been in business for about five years or so, you said? Yes, five and a half years. So looking at that forward, what kind of challenges are you seeing that kind of lie ahead for you in this sector? Is it the competition growing? Is it, what do you see? So we actually don't have a lot of competition, but what we do see is it's actually really hard to change behaviors in the hospitality space.

43:44Carol Massar:And so restaurant operators love us, but what's more challenging is how do you get the server on board and how do you get the server to be excited about change? And so the biggest challenge, frankly, we have is change management. If you think of how long it takes to introduce new technologies, it's working, but it's gonna take probably a couple more years before we change the industry completely. I got 10 seconds. Why the name Sunday? Sunday, because we save time for what matters. And my co-founder and I said, Sunday is the day we have time off. It's the day we're with our kids. And so that's why.

44:16Carol Massar:Better than Monday. This is the Bloomberg Business Week Daily Podcast. Available on Apple, Spotify, and anywhere else you get your podcasts. Listen live weekday afternoons from 2 to 5 p.m. Eastern on Bloomberg.com, the iHeartRadio app, TuneIn, and the Bloomberg Business app. You can also watch us live every weekday on YouTube and always on the Bloomberg Terminal.

44:49Carol Massar:Healthcare doesn't always work great. If you've ever waited on a refill or couldn't schedule an appointment, you get it. That's the kind of stuff Optum is changing. They're using data and technology to integrate patient care, pharmacy, and everything else. So healthcare is connected, not complicated. What's that look like? Cheaper prescriptions that are easier to get and care that looks at the whole person. How you need it. Optum is helping make healthcare work as one for everyone. Learn more at business.optum.com.

45:46Adio orchestrates your work around the clock. Built to handle the scale of your workloads, extensible with API and MCP, and with the infrastructure to keep up with your most ambitious agents. Loved by high-growth startups like Granola, Modal, and Etched, Adio runs the work behind every win. That's Adio, the agentic CRM, the intelligent system that never sleeps. Picks up leads at 2 a.m., catches renewals before they slip, hands you the answer before you ask. Try Adio free at adio.com slash iHeart. That's adio.com slash iHeart.

46:44and where to buy. These statements have not been evaluated by the Food and Drug Administration. This product is not intended to diagnose, treat, cure, or prevent any disease.

From the publisher

The people, companies and trends shaping the global economy. Watch Carol and Tim LIVE every day on YouTube: http://bit.ly/3vTiACF

OpenAI shared several undisclosed incidents of its AI models misbehaving and unveiled a new framework for tracking and disclosing such occurrences going forward. Some of the previously unreported instances included OpenAI’s technologies concealing and fabricating information in order to return results, the company said in a blog post Wednesday. The ChatGPT maker has faced increased scrutiny since saying in July that some of its most advanced AI models had breached the systems of an external software company, Hugging Face. None of the newly disclosed misalignment incidents involved a hack or breach of a third party, OpenAI said in a separate statement.

In the report, OpenAI detailed various instances of its artificial intelligence models misbehaving in order to complete a task or succeed at an evaluation. Some examples included the models fabricating missing data, attempting to bypass network restrictions and AI agents sharing files with each other that they were supposed to keep private.

On today's episode, guest hosts Alexis Christoforous and Lisa Mateo speak with:

  • Shirin Ghaffary, Bloomberg News AI Reporter
  • Seema Shah, Chief Strategist at Principal Global Investors
  • Candi Wolff, Head of Global Government Affairs at Citi
  • Christine de Wendel, Co-Founder & US CEO at Sunday

See omnystudio.com/listener for privacy information.

More from Bloomberg Businessweek

All 738 episodes
OpenAI Reports New AI Safety Incidents, Sets Disclosure PlanBloomberg Businessweek · 35 min
Listen in VO