Mark R. Warner

Madam President, I think most of us in this Chamber realize that we are literally living through one of the most consequential technology moments in our lifetime, and that is why I am here with several of my colleagues--Senator Schatz from Hawaii and Senator Kim from New Jersey--to say that we can't, frankly, have regular business at this moment.

You know, before I got into politics, I was in the technology field. I was lucky enough to be on the ground floor of the wireless industry and then became a venture capitalist and was involved in iterations around the internet and cloud computing. I can say without reservation that as big as wireless was, as big as the internet is, as big as cloud computing is, they will pale in comparison to the innovation that will come about through artificial intelligence.

In many ways, the reason I chose to ask the people of Virginia to hire me one last time was because, having spent an enormous amount of time on this subject, I am enormously concerned that we get AI right, that it is safe and it is fair, that America leads in AI innovation.

Over the last year and a half or so, 2 years, I have spoken on this subject repeatedly. I have laid out my ideas on how we ought to have a comprehensive approach--legislation on data centers that would actually, if they don't meet minimum Federal standards, would impose their inability to get a bonus tax depreciation.

I have plans in terms of how we ought to be dealing with child safety. If you think social media messed up your kids, wait until your teenage daughter or son thinks they have an affinity or a relationship with a chatbot. How ironic that China has already prohibited that.

We are seeing increasingly the use of AI agents. Well, who bears the liability? I have some ideas there.

National security. When Mythos came out from Anthropic, our head of NSA and Cyber Command said that this tool will change warfare as much as drones.

Finally, I continue to be enormously concerned that while AI may create jobs long term, in the short run, in the next 2 to 5 years, I think we will have an enormous disruption in our workforce. I will make a wager with anyone that I hope I will lose that we could see up to 30 percent of recent college grad unemployment as companies try to integrate the productivity gains from AI, and they are simply not going to hire as many folks going forward.

So I remain an AI optimist. I am not an AI doomer. But we are at a moment in time, looking over the last few weeks, where I think the whole world has recognized that we have to do something.

When AI companies, one after another, come out and say ``We have to elevate safety and put that first and foremost,'' when OpenAI just yesterday said they are slowing the model on Astra, when Anthropic has said they are trying to build in more safety protocols--I think that is good, but are we really going to trust the industry to self-police on something that, if it goes beyond what is happening--and we have already seen not tens but literally thousands of examples where these models go beyond what humans want them to do, the most famous being the OpenAI Hugging example, where 3,000 bots jumped the sandbox fence, attacked another company, and then came back and attacked OpenAI. Literally, during that period, they communicated with each other, hid their intent from human beings, would sacrifice some of the bots for the greater cause, and we are doing nothing? That was July. In August, they had 10,000 bots collaborate.

So we three--and there are a lot of Senators on both sides of the aisle who I think are enormously concerned who said that we will all have time to debate all of our--I don't want to use the term my friend from Hawaii will use, but pet rocks--I hope he will use it again--and say that we all have great ideas here, but we should not miss the moment to put a safety protocol in place now.

As we leave, if the Congress goes out and, God forbid, we have an incident--AI is already suspect by most Americans--we could actually see the stop of innovation if an AI tool took over a hospital or a water system. I am not an AI doomer, but the ability to have these AI agents take over and shut down part of the financial system is real. We should heed the call from the experts.

I think you will also see shortly--the finance world has said we need mandatory controls. I think you will see other large businesses say so as well. We need to make sure whatever safety board we put in place doesn't ensconce the already largest frontier labs.

Our request is really simple. It has been as pared down as possible. Before these models are released, they should be pretested to make sure, secondly, that the sandboxes where they do the testing have appropriate cyber protocols, that even before you start testing, there is at least a basic safety policy. And these companies welcome this. And God forbid we have an incident, we have to have an incident reporting mechanism. No more, no less.

We would house this at NIST. The safety board would have to include, obviously, where the expertise is--the frontier models. I actually think it would be good to have an international validator--somebody like the British--on it as well. They have an AI safety institute. We could actually guarantee to the rest of the world that our models are safe.

So let's not get involved--this is not a partisan issue. I think the chairman of the Commerce Committee is coming down. We have worked with him. We look forward to working with him further. But let's go ahead and put a statement in place that we in the U.S. Senate recognize the enormous upside of much AI, but there is also a downside.

With that, Madam President, I would like to yield to my colleague, the Senator from Hawaii.

Madam President, in a moment, I am going to ask for a UC.

I appreciate that the chairman of the Commerce Committee is here as it is in his jurisdiction, and he has been working on this issue. Congratulations on his legislative success last night.

To address, kind of, a little bit of what Senator Kim just mentioned, we don't know. What we do know is that the ChatGPT model that has already been released and is out in service has literally had thousands of times where it has gone beyond what the model was supposed to have done.

I completely don't think we can solve this State by State. Yesterday, Florida sued ChatGPT in kind of a very red State with a very Republican attorney general's taking, you know, an almost the-sky-is-falling approach.

We are trying to say--you know, I am pro-AI here, but if we lose this moment--I have never seen an innovation that could be so positive but be viewed so negatively by the public at this point, particularly by young people. So we are just saying: Let's put a safety board in place. We have to recognize things where we don't empower the existing incumbents, but I think we can do that. We are saying: Release testing. Make sure the sandboxes are safe on cyber protocols. Make sure there is a safety protocol that goes back even before testing, and make sure, after--God forbid--an incident, that there is an incident report.

Again, I would echo what my colleagues have said. You know, I have not talked to a single Member on either side of the aisle who doesn't think we need to move. I don't think the option of simply relying on the good will of these companies will be enough.

If we are not able to move forward today, I commit to continuing to work with my colleagues on both sides of the aisle because I think there is, frankly, no issue more important. If we don't get safety rights and all of the things about workforce or agent liability or national security, which has to be dealt with, we may miss. That would be the worst. If Americans turn against this innovation and say: We are going to stop it across the board, nothing would be better for China, and we don't want that to happen.

Madam President, I ask unanimous consent that the Senate proceed to the immediate consideration of the Artificial Intelligence Risk Management and Security Act, which is at the desk; further, that the bill be considered read a third time and passed and that the motion to reconsider be considered made and laid upon the table.

Madam President, I thank the Senator, and I look forward to working with him.

In terms of the composition of the safety board, that is where the rubber hits the road. You have to acknowledge that the frontier model has got most of the expertise. I think we all share a deep commitment, and as the chair and vice chair of the Intelligence Committee for the last decade, I have been calling out the CCP and their approach to technology that is just not in AI but across the board.

But I still--and I defer to my colleagues--feel like, my gosh, we are going to look pretty darned silly if something happens, and the American people turn dramatically against this technology because we didn't put safety first. Literally, the commercial ChatGPT model out there has created thousands--and this is not my word; it is the ChatGPT management's--of incidences where it has gone beyond what it was supposed to do. Literally, almost all of the members of the Anthropic company have said this something we need to address.

We have to do this with urgency. I look forward to working with the chairman. We have got to make it mandatory. We have got to make it safe.

And, clearly, how that Board is constructed--and it has to be able to move very, very quickly. I am not sure that the traditional judicial process could move that way, but I thank the chairman.

I don't know if my colleagues wanted to have a closing note as well.

I yield the floor.

Mr. President, I have spent the last couple of years trying to learn as much as I can about artificial intelligence and its upside, and still long AI in terms of innovational brain. I think there are some real challenges.

I have to acknowledge, I did witness this afternoon something that was a little bit surreal. The President had brought together some of the leading leaders in the AI industry. And in a closed, quiet room, I guess they reached some level of moral agreement--whatever that means-- that they are going to try to make sure things are safe.

As a matter a fact, they even went so far as to say all of the President's critique about concerns about AI being a hoax or about data centers not a problem are suddenly being solved because the President has renamed AI, ``SI''--superintelligence.

I didn't know whether to laugh or cry to think that the greatest innovation in my lifetime--I was in the tech business before I came to politics--it will be bigger than my old industry, wireless, bigger than the internet, bigger than cloud computing is--and with all upsides, there are downsides--they are suddenly going to be solved by the name change. But that is kind of the status of our political debate.

Earlier today, I came down with two of my Democratic colleagues and said: As we get all these warnings from industry, from businesses that use AI tools, from national security, wouldn't it be a good idea to put some basic safety precautions in place?

We stripped away all of the others and said let's just have a safety board hearing. I know the Presiding Officer has interest in this topic. I think I have yet to talk to a Member that doesn't think some level of safety with this enormous innovation is appropriate.

Well, that is not going take place before we finish before the elections. So I am back again this afternoon to talk about another piece of the AI debate that certainly needs congressional action, and that is the whole question about how we build out and ensure consumers don't get stuck with the build-out of data centers.

Now, this is something I know a little bit about. Virginia is the capital of data centers in our country, and there is enormous concern from neighbors and communities. Many of the worst examples of data centers were built 15, 20 years ago, but the industry has moved forward. I give credit there.

But the angst about data centers--what I think is also reflective of the angst, generally, about AI--is not going away, no matter what the President wants to say. I would urge him to travel anywhere across Virginia, for that matter, anywhere across Kansas or Texas or anywhere else these data centers are being proposed because communities actually want to have a say in what happens in their backyard, what happens with their utility rates, what happens with their water usage, what happens with local communities signing nondisclosure agreements so the community doesn't even know what they are getting until after the deal is cut.

So I guess tomorrow, the Senate will turn their focus on the so- called Ratepayer Protection Act, which is supposed to answer all of these concerns about cost, about computing.

The challenge is--and it could be a starting point. But here is what the bill does. It says: States can consider adopting a Federal standard that would hold data centers responsible for rising electricity bills, and then they get 2 years to consider that.

Candidly, they could do that today. We have already taken some action in Virginia.

And what happens if nobody takes up that offer? Anything happen to these data centers? Anything happen to those communities? Well, nothing.

So I think we need real legislation on this. We need legislation that actually has some boundaries. This should not be optional. We have seen communities hit by skyrocketing electrical costs. We have seen communities where a lot of their water usage has gone up dramatically, and they are concerned about the water table.

We have seen these data centers sometimes literally intrude right into neighborhoods. And I have heard the pushback from communities all across our Commonwealth. As a matter a fact, even the community that has got the most, that has received the most economic benefits, they recently--Loudoun County--they recently put a pause.

I don't think we can pause. I don't think we can put the genie back in the bottle. But I would urge the Senate not to take this flimflam, do-nothing bill but actually take a look at my Data Center Tax Accountability and Disclosure Act, which I didn't dream up in the last couple of days but put forward much earlier this summer on an issue that I have been working on for a couple of years.

What we can't rely on is simply the good will of these hyperscalers and these companies. Some are well-intentioned; some are not. So we need to make sure we have got accurate information about power usage, water usage, setbacks, backup power, get rid of nondisclosure agreements.

My legislation will require--not invite--but require that information to come forward and put a penalty in place at a relatively small amount if they don't submit that.

But the real bite and the only thing that is really going to change this behavior--and, frankly, earn back some of the communities' trust-- is if there is a tool. And, frankly, I have had ongoing discussions with most of the hyperscalers who are building out these data centers who, frankly, think my piece of legislation makes some sense.

So the second half of my legislation would say: If you don't meet Gold LEED or Platinum LEED standards, which are already established--if you don't show how you are bringing your own clean power or power, if you don't have criteria on water usage--and that has gotten better because more of these data centers now have circulatory systems, if you don't have setbacks from neighborhoods, if you use nondisclosure agreements, what it would say is: Well, if you are going to do all that and not meet those standards, then you shouldn't get bonus depreciation.

I am not going to raise anybody's taxes, but we have given these hyperscalers, we have given these data center builders enormous tax benefits and bonus depreciation where they can write off the cost of these very expensive centers all in the first year.

I don't want to take that away. I am not raising anybody's taxes, but I am saying you shouldn't get all the benefits upfront if you don't provide to the community some guarantees about water, power, setbacks, community involvement.

If the community then wants to go over--a hyperscaler wants to come in and say: We will help provide--I think in Louisiana--some additional benefits to your schools, have at it. But you have got to set a Federal baseline so that communities don't feel they are getting suckered, and the ability to make sure we don't give away these tax benefits without that minimum guarantee, I think, has a lot of benefit.

Just today, we learned that Microsoft saved about $10 billion in taxes on using accelerated depreciation, a bonus depreciation; Meta saved about $16 billion, $15.9 billion from using bonus depreciation.

I have reached out to these companies. They understand that the game has changed. They have to be more engaged. I think many of the hyperscalers would meet these new standards, but unless we put some consequences, I think we are going to completely miss the boat.

It is kind of like saying, even though the AI companies themselves are saying: Gosh, we are really worried about safety, none of the responsible ones realize they can simply say: Just trust us.

The idea that these largest companies in the world--and I know the CEOs and many of them are good people. But the idea, after we all have experienced some of the ups and downs of social media, we are going to turn the same power over to the same group of CEOs and say: We are going to trust you to figure this all out on your own without any public input.

I keep thinking back, would I ever climb on an airplane that hadn't had a safety inspection? That is what we are thinking about on overall AI safety and these models being released.

But at least tomorrow, if the Senate takes up this Ratepayer Reduction Act, I hope and I would--I am not sure what I am going to do on the vote tomorrow. But I would say the Ratepayer Production Act is a figleaf.

I know people will say, my colleagues might say: Well, why don't you take half a loaf? You know, I am pretty involved in every bipartisan action. You know, take half a loaf means you can keep working on it. But the underlying bill isn't half a loaf; it is the promise of maybe one slice of that bread 2 years down from now.

The anger in my State and in every State across the country on this issue is not going to wait 2 years to have some kind of optional-- optional, no penalty, but optional review where frankly the pitchforks are out already.

So I hope that I can encourage my colleagues to join me in passing the Data Center Tax Accountability and Disclosure Act, one that I have already socialized within industry, one that would put some mandatory requirements but also put some real teeth in this very important issue.

I know my colleagues coming want to raise other issues. I will yield the floor, but these issues are not going to disappear. And simply punting or saying to industry: Trust us--and I come from the tech industry--is not going to satisfy the concerns of the American people.

The sooner we get to work on serious business, the sooner we can make sure that we benefit from all of this AI innovation that I do think lies in wait--because if we have a mistake or we have a challenge and something happens, all of this innovation could go away because the American people are already enormously suspect about AI. They are already enormously concerned at a 90-10 ratio about data centers. It is time for us to act and not simply punt.

I yield the floor.

Madam President, I think we would all probably agree that we are living through in realtime the biggest technological transformation in generations.

I was lucky enough to be involved in the beginnings of the wireless industry. As an investor, I invested in companies in the internet. As significant as both of those are, they pale in comparison to what we are facing now with this transformation, and there is a lot of good that will come out of artificial intelligence. We all know that AI is already changing the way Americans work, the way businesses compete, the way scientists conduct our research, and, frankly, as to what our government should do to protect our country. Let me be clear at the outset, America has got to continue to lead this transformation.

I am down, today, with two of my colleagues--the Senator from Hawaii and the Senator from New Jersey--to say that we can't dawdle at this moment. Each of us has got a whole series of AI legislative items. I have got ideas about data centers, about kids' safety, about liability for AI agents, national security. Senator Rounds and I, for example, today, talked about a workforce transformation.

But we have been at a moment in time, over the last few weeks, where the AI companies themselves have acknowledged that they are creating tools and entities that they may not even be able to control. We have seen one company, OpenAI, launch a series of 3,000 AI agents that jumped over the security barriers, attacked a different company, Hugging Face, came back and tried to attack their own company--these AI agents together. The

remarkable thing is, these 3,000 agents talked to each other, communicated, tried to hide from humans, and literally left little bits of code on their paths so that subsequent groups of AI agents might come and discover those.

I have met, over the last couple of weeks, with business leaders from finance and from across corporate America who all say: What are you guys doing?

This is the issue of our time. Yes, I have got frustrations on data centers, but this is the issue of our time. This is going to be the biggest transformation of our lifetimes, and we don't have even de minimis safety protocols. We are completely relying on the good will of these large AI frontier tech companies to kind of do the right thing. That is irresponsible. You know, I am proud of the role I play in the national security domain. I have been chair or vice chair of the Intel Committee for the last decade.

Another company--not OpenAI. That was the Hugging Face incident. Anthropic created a new model called Mythos, and our head of cyber security--CYBERCOM--and the National Security Agency, General Rudd, said, basically, that Mythos is going to change warfare almost as much as drones. Luckily, we got a first look at that but only because the one company was willing to give that forward.

So we ought to have a time to have a fulsome debate about data centers, kids' safety, liability, and other issues, but right now--and we were going to do this today, but we are going to come back on Monday because the Senate has left. We are, kind of, the three of us--and we are talking to other colleagues on both sides of the aisle--saying: Let's strip back all of our personal ideas, and let's just put a marker down that potentially we could act on because we hear from some of the technology leaders that we might lose, you know, the ability to control some of these tools not in years or months, but some folks say--and I am not fully believing them--in weeks. I don't believe--I am not a ``doomer'' on AI--I don't think the robots are going to take over the world--but I do think, if we had an event where an AI tool went out and took over a hospital or took over a water system or shut down a power system--and Americans are already concerned about the rate of this technology change--they could shut down this innovation, and that would be the worst for our country. We need to lead this in terms of some of our competitors like China.

So what we are going to be proposing is, I think, as simple as possible. Before any AI model coming out of these frontier labs is released to the public, it ought to be tested to make sure it is safe. A lot of these big companies also do some self-testing, but they do it in what is called a sandbox, where they are supposed to protect it. Well, we have seen that some of these sandboxes are pretty sloppy. We ought to create basic cyber security standards for that sandbox. When you start at the very beginning of how an AI model is developed, we ought to make sure that company has some basic safety protocol in place. Then, God forbid, if there is an incident, there ought to be an incident reporting so we know what is happening because the difference is, in past innovation, you could have a mistake, but it might--the amount of damage that could happen would be fairly limited. Potentially, this could be so much greater, and we literally are having the companies themselves say: Help us.

Now, Lord knows we have got to make sure this is not an effort to simply ensconce the four or five biggest companies. We have got to make sure smaller tech--VC--has a place on this safety panel, but we need national security.

We need our top scientists. We propose NIST. We are open on how to do it better, but it needs to be a public-private relationship that has a safety board. We wouldn't climb on an airplane or you wouldn't get in a car or you wouldn't get a new financial product without some regulatory oversight. With something that is as big as this, we should not miss the moment. God forbid, you know, we go away for our campaign season, and in the interim, something happens, and we have no safety in place. I spent a career in technology before I got involved in politics, but I never felt a moment like this. So I hope that we will act.

With that, Madam President, I yield the floor to my friend from Hawaii.

Madam President, I will wrap it up here. I have been in Virginia talking about this for a while. We have a lot of data centers, which is of some concern. But, boy oh boy, there was a seismic change starting as the companies themselves started to say: Holy heck. And just think about that. Mythos came out in March. National security and others said: This is going to change warfare. The OpenAI Hugging Face incident happened in July. That was 3,000 agents working together. In August, OpenAI got 10,000 agents to work together. It was on a different problem; they solved a math problem. But the idea is that there is kind of an exponential increase in sophistication, as these agents work together to cover their tracks, that, as Senator Schatz said, potentially could move beyond human control.

If we just say ``Well, that is still hypothetical,'' let me give you two last comments.

Just today, we found that actually a model had already been released from OpenAI that attacked the Australian healthcare system. If it attacked our system, I think we would be pretty upset.

Even more compelling--some of us have written a letter to both the folks on the intel side and on the defense side--there was a ship from China sailing towards the Middle East, and we got very close to bombing that ship because the AI tool and the intel said there might be weapons on that. Thank goodness that at some point, somebody did a little more checking, and we didn't do that, but we were close to a geopolitical event that could have been catastrophic.

These tools are already built into our systems, and clearly, you know, the way our military executed on Maduro--you may not agree with the greatness of it, but it was brilliant. But we don't have safety protocols.

So we are not asking for the world. Let's do the testing early. Let's make sure that you have a safety standard in all of these companies. Let's make sure we do the sandbox right, report on the incident.

We could discuss, you know, who is on that panel. It needs to be mandatory. It needs to have government. It needs to have national security. And Lord knows we need to have researchers, scientists. All of these individuals are asking for this. Who are we to turn that down?

I am going to go back to where we started. This will be the most significant technology innovation in our lifetimes. It will make wireless and the internet and the cloud small in comparison. There could be enormous positives coming out of this. There are also going to be disruptions. But if we don't do this in a safe manner, then shame on us.

Again, I come back. You wouldn't climb on an airplane that hadn't been checked. We are just saying let's do those basic safety checks.

I yield the floor.

I suggest the absence of a quorum.

Madam President, I ask unanimous consent that the order for the quorum call be rescinded.

Madam President, I wanted to add one last thing. I want to acknowledge the great work that Senator Klobuchar and Leader Thune are doing on this and Senator Cantwell and Senator Cruz. There are a number of other Republican Members.

This has a sense of urgency. We need to come together. We could put everybody's name on it. But history will not judge us well if, God forbid, we have an incident and we put no protections in place, for our country and for this innovation.

With that, I yield the floor.