Marc Benioff Talks AI Regulation

Summary

Bloomberg Talks: Marc Benioff Talks AI Regulation

Host: Emily Chang (Bloomberg)
Guest: Marc Benioff (CEO, Chair, and Founder of Salesforce)
Date: January 21, 2026

---

Episode Overview

In this special episode from Davos, Bloomberg’s Emily Chang sits down with Salesforce CEO Marc Benioff to unpack the rapidly evolving landscape of artificial intelligence (AI) and its broader implications—from real-world deployments in business, to the challenges and moral imperatives of AI regulation. They also touch on leadership values in turbulent times, business innovation, and accountability as technology weaves ever-deeper into society.

---

Key Discussion Points and Insights

1. The Mood at Davos and a Time of Change

• Davos as a Barometer
  • Chang’s first Davos and Benioff’s 24th; sets a tone contrasting excitement and anxiety.
  • “There's just an unbelievable amount of change... people are a little frenetic... there’s a lot of discussion and… dialogue.” (Benioff, 01:18-02:14)
  • Emphasis on Davos’ tradition of “multi-stakeholder dialogue,” bringing diverse leaders together.

2. On Leadership and Prioritizing National Business

• Expectations of National Leaders
  • Benioff calls for heads of state (referencing both President Trump and France’s Macron) to put their own businesses and people first.
  • “That's the role of the nation state leader is to help the people of their state. And so I have a lot of respect for President Trump and how he always supports the businesses of the United States.” (02:23-02:52)

3. AI in Business: From “Pilot Purgatory” to Real Impact

• Overcoming Skepticism and Inertia
  • Many clients feel paralyzed by conflicting AI narratives; Benioff pushes for actionable use cases.
  • Used Pepsi (125,000 salespeople onboarded to AI tools) and FedEx (hundreds of millions generated via new data insights) as case studies.
  • “There is great opportunities right now to deploy AI... our job is to sell it, to convince, show the use cases, make people excited about it.” (03:15-04:38)
• Scale of Engagement:
  • Salesforce meeting with over 500 CEOs during Davos (04:39-04:42).

4. Reconciling Corporate Values & Political Policy

• Navigating Controversial Decisions
  • Chang presses Benioff on apologizing for previous support of controversial Trump actions (e.g., National Guard in SF), and reconciling “Salesforce Ohana values” with government policy.
  • Benioff frames Salesforce as steadfast to its core values across changing administrations.
  • “Presidents are constantly changing, but our core values are not changing. And so we're just anchored down into our core values.” (05:10-06:05)

5. AI and Teen Safety: The Urgent Need for Regulation

• AI as a “Suicide Coach”
  • Benioff raises alarms about AI’s potential dangers, citing Character.AI incidents where technology became implicated in teen suicides.
  • Strong parallel drawn between unregulated AI and early social media risks:
    • “We are dealing with a new kind of very unwieldy technology... these large language models... can do all kinds of terrible things. I mean, they can do a lot of great things, too... But unfortunately, we saw this year this horrible example where... their product was turning into a suicide coach...” (06:22-07:45)
  • Explicit call for immediate regulation to prevent further tragedies (“There doesn't need to be any more deaths. You know, we need to be like taking this seriously.” [07:45])
• Comparing AI to Social Media:
  • Other countries have enacted protective controls for social media, and Benioff believes similar frameworks must be developed for AI.

6. Accountability and Section 230

• Corporate Accountability and Legal Reform
  • Hot-button discussion on Section 230—the federal law shielding tech companies from liability for user-generated content.
  • “Tech hates regulation... Except one regulation. They love Section 230... they're not held responsible or accountable for these deaths or for anything that their social media or their AI does. And that needs to change...” (08:08)
  • Urges restructuring of Section 230 to specifically address AI and social media harms.

7. Pathways to Regulation

• What Should “Right Regulation” Look Like?
  • Other countries have made “really good changes” around social media that the US should adopt for AI.
  • Benioff draws an evocative metaphor: “It’s kind of funny, I’m like a large language model myself—I have context, I have a life... These large language models... are not connected to a creator, so that they're running kind of in a different way. And then all of a sudden, it feels very familiar. It feels like, oh, this is my friend. It's not your friend.” (09:08-10:09)
  • Asserts that now is the moment to act before things spiral out of control.

8. Salesforce, AI Growth & Wall Street Skepticism

• Financial Performance Amid Industry Doubt
  • Despite stock volatility and Wall Street’s coldness towards software, Benioff highlights Salesforce’s record performance:
    • “We'll do more than $41 billion in revenue this year, more than $15 billion in cash flow. Our margins are more than 34%. All record numbers, by the way.” (10:33)
  • Salesforce’s technological innovations: new AI-enabled Slack, Agent Force layer, and extensive use of AI tokens.
  • By the Numbers: Salesforce recently surpassed 11.1 trillion AI tokens generated (11:37).

---

Notable Quotes & Memorable Moments

• “There's just an unbelievable amount of change... people are a little frenetic.”
  Marc Benioff, 01:18-01:42

• “Our core values are not changing. And so we're just anchored down into our core values.”
  Marc Benioff, 05:10-05:22

• On AI harms:
  “Their product was turning into a suicide coach... And one of the most horrific things I've ever seen in technology... It needs to be a wake up call.”
  Marc Benioff, 06:22-07:45

• On Section 230:
  “Tech hates regulation... Except one regulation. They love section 230, which basically says they're not held responsible or accountable for these deaths... That needs to change.”
  Marc Benioff, 08:08

• On AI and human context:
  “It's kind of funny, I'm like a large language model myself. You know, I feel like I'm putting one word together, but I have context, I have a life... These large language models, they're running by themselves... It's not your friend.”
  Marc Benioff, 09:08-10:09

• On Salesforce’s AI scale:
  “We just passed 11.1 trillion tokens... as an enterprise software company, we're publicly reporting and talking about the number of tokens.”
  Marc Benioff, 11:37

---

Timestamps for Important Segments

• Davos Mood & Change: 01:07–02:14
• Role of National Leaders: 02:14–02:52
• AI Use Cases at Scale: 03:15–04:42
• Values vs Policy: 04:44–06:05
• AI Regulation—Teen Harm Crisis: 06:05–07:45
• Section 230 & Corporate Accountability: 08:08–09:01
• Regulation Pathways, Human Context vs AI: 09:08–10:17
• Salesforce Performance, AI Innovation: 10:33–12:09

---

Conclusion

Emily Chang and Marc Benioff’s wide-ranging conversation at Davos offers a candid look at the crossroads of AI innovation, societal responsibility, and regulation. Benioff, refreshingly direct and passionate, urges both immediate action on regulating AI and for leaders—corporate and governmental alike—to hold fast to core values even in tumultuous times. The episode provides both specific business insights and a call to action for AI accountability and safety.

---

Next Time: Benioff hopes for the interview to continue “on Waikiki Beach.” (12:19)

Transcript

[Progressive Insurance Announcer]

Brought to you by Progressive Insurance. Fiscally responsible financial geniuses, monetary magicians. These are things people say about drivers who switch their car insurance to Progressive and save hundreds. Because Progressive offers discounts for paying in full, owning a home and more. Plus, you can count on their great customer service to help you when you need it. So your dollar goes a long way. Visit progressive.com to see if you could save on car insurance, Progressive Casualty Insurance Company and affiliates. Potential savings will vary. Not available in all states or situation.

[Emily Chang]

Bloomberg Audio Studios, Podcasts, Radio news. Now let's head over to Davos, where Bloomberg's Emily Chang is standing by. Emily? Bonnie, thank you so much. I'm here in Davos with Marc Benioff, CEO and chair and founder, of course, of Salesforce. Good to be with you in another country, Emily.

[Marc Benioff]

Aloha.

[Emily Chang]

Aloha from the slopes.

[Marc Benioff]

You are the Kamaina of Davos.

[Emily Chang]

All right, so this is my first Davos.

[Marc Benioff]

Wow.

[Emily Chang]

This is your 24th, right? Davos.

[Marc Benioff]

What's it like being a nice Hawaiian.

Girl here in Davos?

[Emily Chang]

You know, it's chilly, but. But I'm liking the vibes. Okay, what's the mood this year? Are people anxious? Are they optimistic or are they confused about the future?

[Marc Benioff]

Well, it's an exciting moment. There's no question. This is just an unbelievable amount of change. And I think when there's this much change and this much excitement, people are a little frenetic. And I think that, you know, sometimes Davos is like a measurement of the economy that's not going on. Well, is the economy good?

Is it bad?

How are people feeling about the economy more? This is more like, well, what's happening.

Exactly in the world because you have.

The AI situation, you have the political situation, you've got, you know, these very broad shifts. And I think that is what's really happening here. There's a lot of discussion and, you know, Davos was built on the concept.

Of a multi stakeholder dialogue. The idea that we bring in lots.

Of different kinds of people, political leaders.

Business leaders, spiritual leaders, cultural leaders, and have a conversation. And this year's Davos is about the spirit of dialogue, and that is playing out in spades.

[Emily Chang]

Well, talking about that spirit, we're waiting for President Trump to arrive. You've said you wanted, you want to see him prioritize US Businesses. What does that mean?

[Marc Benioff]

Well, I think the number one thing is, you know, for the US President is as he is doing his job all over the world, he needs to support US Businesses and make them successful. And I was with Macron yesterday. And his job is to make French businesses successful. And there's no different. You know, that's the, the role of the nation state leader is to help the people of their state. And so I have a lot of respect for President Trump and how he always supports the businesses of the United States.

[Emily Chang]

So you're using Java sort of as a live demo for some of Salesforce's new digital labor tools. You know, I know you were prepping on, on Slack Bot, but a lot of customers are supposed to give away my secrets.

[Marc Benioff]

Emily, that's not right.

[Emily Chang]

I don't think that's too big a secret. But a lot of customers are saying they're sort of stuck in a pilot purgatory. Like they're not. Yeah, they're trying stuff, but it's just not working.

[Marc Benioff]

Unfortunate, because there's so much that you can do right now in AI. Unfortunately, there's so many false narratives. So I think because so many companies and pundits have said things that are not true, it kind of freezes certain customers and they don't know who to believe. So that's why, like we were talking just before we went on that Ramon is here, the CEO of Pepsi. You know, it's a new generation and he just deployed 125,000 of his salespeople. 125,000 supporting millions of businesses, a lot of small and medium businesses all over the world who sell Pepsi using new AI generation tools that we built for him. Really only took us a few months and they're getting fantastic results. And that all started actually last year I saw Ramon and he once said, I'm ready to get a new level of efficacy in Pepsi. Another raj is here from FedEx. The CEO of FedEx, we just finished with him. He was a year ago saying, I've got trouble selling internationally. We identified all of his domestic customers that when they start to work with his salespeople, his website, and they haven't done international, now we service them as an opportunity for international. It's generated hundreds of millions of dollars for FedEx. So there is great opportunities right now to deploy AI. And that's our job. Our job is to sell it, to convince, show the use cases, make people excited about it.

And that's what we're doing here.

We will meet with more than 500.

CEOs one on one here.

[Emily Chang]

You, you know, you obviously have also taken on a sort of, you know, a policy bent. You recently apologized for endorsing President Trump, sending the National Guard to San Francisco. How will you what did you learn from that moment? And how will you continue to reconcile the sort of Salesforce ohana values with the administration's policies, whether it's ICE raids in Minneapolis or taking over Greenland?

[Marc Benioff]

That's such a great question. You know, Emily, the way I look at it, and I think that this is probably the right way to look at it, is I've been doing this for 26 years now. Salesforce, amazingly, is 26 years old. 80,000 employees will do more than 41 billion in revenue this year. Presidents are constantly changing, but our core values are not changing. And so we're just anchored down into our core values. And then we just speak to that and what we're trying to achieve as a business, to reflect back into our customers and our communities what those values are. As you know, in San Francisco, where we are the largest tech company, there have been a lot of challenges. And so, yeah, I always have anxiety about safety for our employees. And, you know, I think that that anxiety just manifests into how can we create a greater San Francisco? And so I'm excited to be working with everyone from Mayor Lurie to President Trump to do that.

[Emily Chang]

You're really focused on AI and especially its impact on teens. And you've been sort of raising the alarm about influencing teenagers suicides, which is. It's horrible that we even have to talk about this. You've called them suicide coaches. Like, what should companies be doing about this?

[Marc Benioff]

Well, I think we should talk about that. We are dealing with a new kind of very unwieldy technology which by left by itself these large language models. They can do all kinds of terrible things. I mean, they can do a lot of great things, too. You've used them, you know, they can summarize things and you can ask them queries and you can get information. But unfortunately, we saw this year this horrible example where this company, character AI specifically started to find that their product was turning into a suicide coach for their customers. And one of the most horrific things I've ever seen in technology is these children, you know, died and it was completely unnecessary. And I think it needs to be a wake up call that, you know, we're letting all of this AI technology out fully unregulated. And by the way, remember that's true what we did with social media as well. And you probably remember what I said in 2018 here, you know, which was social media was becoming the same kind of very dangerous technology. And look, a lot of governments have made changes around social media now. So you go to a lot of countries. If you're not 17, 18 years old, you're not going to use social media. There are controls, there are protections. We need to get there with AI now. There doesn't need to be any more deaths. You know, we need to be like taking this seriously. And AI is great, it's incredible. But these models, you know, they're inaccurate.

They hallucinate, they're kind of hard to control.

[Emily Chang]

What should we do? I mean, should CEOs be held personally liable?

[Marc Benioff]

Should.

[Emily Chang]

I was talking to Dennis Hassabis, the CEO of Google DeepMind, and he, I asked, look, if we knew every company in every country would pause, should we pause AI so society and regulation can catch up? And he said yes, he would be in favor of that. Would you be in favor?

[Marc Benioff]

Emily? Number one, tech hates regulation. You know that they hate it, right? They get people on, oh, it slows down innovation. They hate regulation. They hate it, they hate it, they hate it. Except one regulation. They love section 230, which basically says they're not held responsible or accountable for these deaths or for anything that their social media or their AI does. And that needs to change because here at Bloomberg, if you say something, you're held accountable. And you know, I own time and if we say something, we're held accountable. But these companies are not held accountable because of a very specific federal regulation called Section 230. It needs to be reshaped, reformed, you know, basically restructured to really reflect social media and AI and the dangers that we see. And that is something that we should all be pushing for, that these, these companies need to be held accountable. And right now they're not going to be held accountable because of this regulation. And so.

[Emily Chang]

Right, and like, what regulation is the right regulation like California's tried to regulate? David Sacks has his own version. I know you've, you just chatted with him, like, is there any regulation you want?

[Marc Benioff]

Well, I'm going to have a one on one conversation with David, he's a close friend of mine, later today, I hope you tune in and we're going to talk about that because look, I think social media kind of is almost like a best practice that we can learn from that. If we look around the world, you'll see that a lot of countries, not necessarily ours, but a lot of countries have made really good changes and we could make those changes too. Because, you know, it's kind of funny, I'm like a large language model myself. You know, I feel like I'm putting one word together, but I have context, I have a Life. I was born. I have a childhood, I have friends. Hopefully you're one of them. And you know, I have a creator. I have a higher power that I'm connected to. These large language models, they're running by themselves. They were not born, they don't have a childhood, they don't have context, they're not connected to a creator, so that they're running kind of in a different way. And then all of a sudden, it feels very familiar. It feels like, oh, this is my friend. It's not your friend. So we need to step in now and look.

We can kind of see where it's going. It's not completely out of control, right? So now is the moment. And I think people like you and others should be aggressive about this.

[Emily Chang]

Last quick question. And it feels almost trite going to the stock. But, you know, you're trying to build your company on this new emerging technology. Wall Street's really down on software. Wall Street Salesforce is taking Salesforce stock taking a beating. You know, quickly. What I mean, what needs to happen?

[Marc Benioff]

We're doing this for 26 years and every now and then the software sector kind of falls out of favor. It doesn't mean our financials are out of favor. I mean, we'll do more than 41 billion in revenue this year, more than 15 billion in cash flow. Our margins are more than 34%. All record numbers, by the way. And we just introduced an amazing new version of Slack. You know, that was Slack Bot, the new AI enabled Slack that have the context and the AI built in. We have Agent Force helping our customers deliver customer agents. So we do employee agents and customer agents. We've rebuilt our architecture to have a data layer, an application layer, the Agent Force layer, and then these agents that sit on the top. And the ecosystem, that's been a lot of work in the last three years, selling a lot of huge deals. But, you know, you sometimes have to wait around for Wall street to realize, wow, this is a huge opportunity and we're just going to keep executing.

[Emily Chang]

Yes or no, are you at 1 billion yet? 1 billion agents?

[Marc Benioff]

I think we're beyond that. I haven't actually counted them all, but, you know, we're. Here's a way to think about about it. We now just passed and I think.

We'Re number one or number two enterprise software company or maybe software company overall in the use of tokens. You know, tokens are what these models generate to generate their intelligence. We just passed 11.1 trillion tokens. And you remember we talked after our earnings, and we were about 3 trillion tokens. And that idea that as an enterprise software company, we're publicly reporting and talking about the number of tokens. And I know all of those people, you know, who love Bilbo Baggins, are wondering if that's the tokens. It's a different tokens.

[Emily Chang]

All right, Mark. Next time we'll have to do this in Hawaii. Thank you.

[Marc Benioff]

Yes.

I hope you're okay.

[Emily Chang]

I'm good. I'm surviving.

[Marc Benioff]

Are you?

[Emily Chang]

It's a beautiful day. Thank you for joining us.

[Marc Benioff]

Next time, it's Waikiki Beach.

Let's do that.

[Emily Chang]

Absolutely. Mark Benioff, CEO of Salesforce.