IBM CEO warns this would ‘NOT BE GOOD’ for US in AI race…

Speaker A [00:00:00]:
Joining us now to talk about the ramifications is IBM chairman and CEO Arvind Krishna, who is at IBM's Think 2026 conference in Boston. How many people are there, Arvind? A lot of people coming to see you.

Speaker B [00:00:14]:
We got almost 10,000 people had a keynote this morning, room was packed, had all our partners here yesterday and room was overflowing. So it's great to see the interest and the excitement.

Speaker A [00:00:24]:
Well, 10,000 people, that's got to be a record. Unbelievable. Arvind, you're revealing all kinds of related innovations at this Think conference that we definitely want to hear about. It's happening at the same time. The Trump administration announced today that Google, Microsoft and Xai have agreed to hand over their new AI models to the government before releasing them to the public so that the government can evaluate the national security implications. How does giving that kind of control to the government sit with somebody like you as the CEO of a company that's constantly innovating?

Speaker B [00:01:01]:
Look, this is always the balance between innovation and safety. As long as they're going to do their judgment quite quickly within a few days or a few weeks, I think that this serves everybody very well. If it turns into a bloated bureaucracy, that would not be so good for us to win the race.

Speaker A [00:01:20]:
That's a great point. I mean, some would cheer government oversight of AI as much needed regulation. But is it a slippery slope where the government might expand the early reviews from frontier AI models to government intervention, I don't know, into innovations IBM is making for enterprise?

Speaker B [00:01:38]:
Look, I think we live in a regulated world and that is good to have the guardrails around innovation. If I look at it, the banking regulators did not say that AI is a different technology. They were always going to regulate the use case of it, whether it's in payments or it's in terms of customer service. In health care, the same thing applies. In telecom, the same thing applies. So there is always a level of government oversight. And I get back to the balance between too many regulations, terrible, too few. We may not love the outcome. So we've got to find the Goldilocks Medal.

Speaker A [00:02:14]:
Well, you found a Goldilocks area in developing the technology that you're unveiling at your show. I want to hear about some of it. Specifically the Watson X Orchestrate. This is for enterprises, right? And what does it do in that is different from anything else that's out there or that you even at IBM have come up with?

Speaker B [00:02:33]:
Yeah. So Liz, we did a number of announcements here. Orchestrate is one and I'll touch on that in just a minute. But IBM Bob, which helps enterprise developers across the software development life cycle concert, that helps them maintain the security of their entire infrastructure, sovereign core, that helps people deploy private clouds in a sovereign way. So to get back to orchestrate as we all want to deploy agents and we actually acknowledge there will be plenty of third party agents, meaning from other sources, but you need to deploy them in a way you can monitor them, you can audit them, you can make sure they're only getting to the data source you want them to get to and they're being used for the purpose you want them used. To orchestrate helps you orchestrate all of your agents, whether they're from the outside, that third party inside, or their own developed agents. And that really gives the Enterprise and the CEOs and the CTOs control over the infrastructure and also gets the real time data because that is how you unlock the value of AI and it makes that available to these agents.

Speaker A [00:03:38]:
There are so many companies out there that are technologically very advanced who are rushing to create all kinds of AI agent opportunities for businesses for enterprise software. So the competition is rough and tumble at this point. I mean think about it, Arvind, you've got Allbirds, a shoe company converting to an infrastructure type of company. I mean I've had people say, wait, what? It's a real land grab here. So how do you at IBM make sure that you're getting your products in front of these companies before the other guys do?

Speaker B [00:04:12]:
We'll win on the basis of our incumbency. The fact that we could bring 10,000 people to this conference and they want to hear about our products. Behind me you can see them interacting with our engineers and our sales teams in learning about the products. That's one advantage for us. Second, people will run. That doesn't mean that they've actually done anything that is of real value to an enterprise. I'll just say it this way, the next few months will separate the wheat from the chaff and we think we are part of the substance and we can help get real value. We had a number of great clients this morning who talked about how we are helping them get AI embedded in their enterprise. We had Elevance Health, who's a health care company, a pair with over 22 million commercial customers in addition to Medicare and Medicaid. They talked about using our help, Aramco, the world's largest energy company. They talked about how we are helping them get going. We had the Cleveland Clinic talk about how we are helping them improve health care outcomes, not just in research but in how they provide them. I think these are the areas where we will tend to win. And in the end, I think quality of engineering, quality of delivery and delighting our clients to make sure that actually improves their business and is not just a shiny little experiment on the side is the key differentiation.

Speaker A [00:05:34]:
But there are some chief investment officers and technical officers who are a little bit resistant to it. I mean we hear from them, they're afraid because they're already spending so much on old school suites of software that are very expensive and heavily embedded. How do you convince those companies to basically cut the umbilical cord and go all in on something like this?

Speaker B [00:05:58]:
I think that the reaction to some of the SaaS pricing changes as well as the seed based pricing, we're going to see that play out Whenever there's a new technology that comes in, be it it was Internet circa 1995 to 2000, now it's AI it takes a little bit from what was already there. What was already there doesn't go away, but probably gets somewhat compressed in shape and then some of that is going to get replaced. This is the point of innovation. We have to run to it because if we don't run to it then what be otherwise have would get compressed. But being an AI allows us to expand. And I think the latest numbers we printed in the first quarter in the month of April kind of tell us that we grew our software revenue by 11%. We grew the company by over 5%. Those are really good numbers. And I think that that speaks to our ability to take share in this emerging paradigm.

Speaker A [00:06:56]:
You guys have not been immune to the panic that swirls around the these headlines that every single time somebody comes out with something new. And I'm thinking anthropic and the Claude code tool because February 23rd that happened. You guys were not immune to it. Your stock got hit because this Claude Code tool basically showed that it could update old computer systems and provide a computer language that would be perhaps. I mean, I see, I know your reaction here, but you got to tell investors and this investment audience that you're talking to at the moment about why something like a new development for Manthropic isn't going to derail IBM.

Speaker B [00:07:36]:
The day that the cloud code blog on COBOL was written, we were down 13%. We got a lot of it back over the next three weeks. So this is going to go in sort of all directions I, I think for the next year or two until people can figure out who are the winners. But at the end of the day, if enterprises get value from our software and our infrastructure and we have revenue growth and we can get market share, then this is to me, short term or mid term noise to some extent. And in the end the real numbers will tell the story. I think the bigger story is that software multiples as a whole have come down and we are certainly not immune from that. I'll acknowledge that. And if those multiples come down, then maybe that's an opportunity for us to be opportunistic in the market.

Speaker A [00:08:25]:
Well, I'll tell you, you got a lot of people, 10,000 of them, waiting to see all the developments at IBM, at the Think Conference in Boston. And one of those 10,000 is Diplo, the DJ. How cool are you guys?

Speaker B [00:08:37]:
It is. And a lot of excitement also on our quantum technologies. We just announced last night that Cleveland Clinic did a lot of great work and they are now analyzing a 12,000 atom protein trips in using quantum computing. I think that this is a real moment where people are doing real biology using quantum computing.

Speaker A [00:09:01]:
Arvind, thank you.