Navigating the Realities of AI Policy: The 2026 Data + AI Policy Intelligence Brief

Summary

This video examines the 'Deregulation Paradox' enterprises face in 2026: while the US federal government is rolling back AI rules, states are simultaneously introducing over 1,200 AI-related bills, creating a fragmented compliance landscape that is difficult to plan for.
Leading experts from Databricks, Mehlman Consulting, and Andreessen Horowitz explore the EU AI Act, development-side versus deployment-side obligations, and where federal preemption is likely versus where state enforcement will prevail.
Practical guidance is provided for enterprise engagement with policymakers, compliance strategy for regulated industries such as healthcare, finance, and critical infrastructure, and how to separate regulatory signal from noise in a rapidly changing environment.

Navigating the Realities of AI Policy: The 2026 Data + AI Policy Intelligence Brief

Artificial Intelligence and how to regulate it is top of mind for companies, governments, and policymakers everywhere. Today, industry faces a “Deregulation Paradox”: While the current US Government takes executive action to dismantle AI rules, states are doubling down on their own policies and enforcement. In 2025, U.S. legislators introduced over 1200 AI-related bills, threatening a regulatory patchwork that is colliding with a federal push for preemption. To separate the signal from noise, join leading experts for a deep dive into the the policy and politics surrounding AI and data regulation at the federal, state, and international levels in the year ahead. Moderator: Alan Davidson, Databricks Head of Government Affairs
Talk By: Alan Davidson, Head of Government Affairs, Databricks ; Bruce Mehlman, Founder, Mehlman Consulting ; Matt Perault, Head of AI Policy, Andreesen Horowitz (A16Z) ;

Chapters

FAQs

What is the Deregulation Paradox for AI in 2026?

The Deregulation Paradox refers to the tension between the US federal government's actions to dismantle federal AI rules and the simultaneous surge of state-level AI legislation, with over 1,200 AI-related bills introduced in 2025. Enterprises face conflicting signals about what compliance obligations they should build for.

What is the difference between development-side and deployment-side AI obligations?

Development-side obligations apply to organizations that build AI models and systems, while deployment-side obligations apply to organizations that deploy and use AI in their products and services. The distinction matters for enterprise compliance planning because a company can face both sets of requirements simultaneously depending on how it engages with AI.

How should enterprises engage with AI policymakers according to this video?

The panelists advise that organizations who choose to be heard in the regulatory process are better positioned than those who wait to be overruled. Practical engagement includes participating in public comment periods, engaging with industry associations, and building direct relationships with government affairs contacts who understand both the technical and regulatory dimensions.

Which industries face the most immediate AI regulatory risk?

Healthcare, financial services, and critical infrastructure are identified as regulated industries facing the most immediate and specific AI compliance requirements. These sectors already have existing regulatory frameworks being extended to cover AI applications, making early compliance investment more pressing than in less regulated industries.

Full transcript

[00:14] Last year, over 1,200 pieces of legislation regulating some aspect of artificial intelligence systems was introduced were introduced in states and legislatures like across America.
[00:26] And many, many more bills were introduced in capitals around the world. Um most of those bills did not pass, did not become law, but it is so clear that governments around the world and people
[00:38] around the world are now asking these questions. What's the right role for a government and for regulation, for AI policy in the middle of this new AI revolution?
[00:51] The answers are totally unclear and they are changing all the time. Just read the papers. Well, I guess people don't read the papers anymore. I guess I would say look at your newsfeed and uh you can see how quickly it's changing. Uh and so for
[01:05] builders and developers and technologists, practitioners, it is a confusing time, right? What rules should we expect? What do we need to build for? How do we best
[01:19] prepare? How do we best engage? Um to help us answer these questions, uh we have two outstanding experts here with us today in technology and AI
[01:31] policy and politics. And uh so I just say welcome to our session on navigating AI policy. I'm Alan Davidson. I'm the head of government affairs at Databricks
[01:43] and I will be your moderator today. Uh I will be your cruise director for this conversation for the next 38 minutes. They're they're counting us down. Um I will just say as way of introduction,
[01:55] the bios for our two panelists are in the program, but I wanted to just say a very quick word or two about why I am so excited that they are here with us. Um
[02:07] Both people I've I've had a chance to admire for some time now. Bruce Mehlman, just to my left, has been really a towering figure and thought leader. I'm going to say that. Uh in tech policy for over 20 years in Washington, D.C. for
[02:20] longer than that. Uh he's a former assistant secretary of commerce. He started Mehlman Consulting, which is a very renowned firm in Washington. He runs a very popular Substack, which I commend to you all called The Age of
[02:32] Disruption. You get six slides every Sunday and it kind of like blows your mind a little bit. Uh he's an advisor to boards and CEOs and we're extremely lucky to have him here today. To his left, Matt Perault is the head of
[02:45] AI policy at Andreessen Horowitz. What a cool job. Um Matt is uh he was also the head of uh policy development at Meta for years. Uh also at Duke, he's a
[02:57] one of the deep thinkers in this space and um a policy wonk's policy wonk. That's what I was going to say and I say that with love and admiration. >> it as a compliment. >> Thank you. Yes, I yeah yeah. Um so uh
[03:09] we've got um yeah, some some of the leading lights here to help us uh navigate this in the next um the next few minutes. So, let me just start by saying, please join me in welcoming our panel our our our expert
[03:21] guests today. >>
[03:33] time from across the spectrum, right? Uh technology companies, healthcare companies, retail, defense, like about what's going on in Washington, about what's going on globally. Um AI is coming of age at a pretty fraught
[03:46] time in the world. Uh help ground us just as a kind of start, like in the broader moment. Like, what is driving politics, policy, voters right
[03:58] now? What is driving the conversation? What's top of mind for people as we start to think about AI policy? >> Well, first, thank you and and thanks everybody for showing up for our panel. Knowing we're competing against pretty cool stuff, but thanks to Data Bricks
[04:10] for the invitation and it's always great to be here. I probably won't fence with Matt, but it'd be fun if we mixed it up about something. There is AI is showing up at
[04:22] at certainly an up a rising. I don't know if we're at the peak or if it's continuing to crescendo of what in that you mentioned my Substack, you mentioned you forgot to mention the most important part, it's free.
[04:34] But but the name of it I I I've named it Age of Disruption because we are in an age of disruption. And it it's really easy when you look at the news, when you watch TV to get really bent around the axle about any
[04:47] given individual political figure or any given moment. But I think where we are is at the amidst a multi-decadal series of changes, technological changes where
[05:00] with the information revolution, the internet becoming AI, everything how we work, live, play, learn, parent, fight wars is radically changing and that's accelerating. And for a lot of
[05:13] people, they don't know if they can keep up and they're feeling scared. They're they don't trust institutions, they're voting for change. While that's happening, the world's changing. So, you go back in 1984, Mac Apple
[05:26] starts selling Macs, Michael Dell starts selling customized PCs out of his dorm room at University of Texas. And we quickly get to this period from 1991 to 1990 4 where the Cold War ends and you
[05:40] have all the former Soviet Union, all those countries join the global economy and India and China with big reforms and you'd have billions of new consumers and billions of new producers. And companies like Databricks suddenly can tap global talent, global capital,
[05:53] global markets. Radical geopolitical change. At the third big change at the same time is culturally. Where America has gone from an overwhelmingly white, Christian, non-college educated
[06:06] country not that many opportunities for women in the workforce to a much more pluralistic society, stronger in so many ways, but we've also atomized our media. We've atomized people now get not the
[06:19] news, they get their news. And a lot of people don't think there is the truth, they have their truth. And so at this time people look at the technological, geopolitical, cultural changes that are happening faster. They
[06:32] don't know how they can keep up. They don't trust the institutions have their back, so they're voting for populists and they're voting for people who they think maybe will deal them in and and and help help them succeed in this
[06:46] environment. So it's both the best of times and the most stressed of times. And a lot depends on who you are, but even folks here who are I would argue clearly in the winner's circle. I
[06:59] bet most people here worry about can we keep up and you know, it's you're you're only you get Databricks is as great as it comes and I know from our dinner last night we're also recognizing there's a lot of big guys
[07:11] who would want to do what we do. >> So that So all right, Matt, help us drill down a little bit. So that's the That's the biggest frame here. I mean, okay, Matt, you've got you're the top tech policy expert at the
[07:24] you know, top venture capital firm on the planet. How do you all look at this moment and especially as it comes to like tech and AI? >> Yeah, so thank thanks a lot, Alan. I started
[07:37] on almost a daily basis after I started I heard from other people at the firm about how much admiration we have for your company, for your leadership, for your products. We are extremely
[07:49] loud, vocal, adamant champions for you and your work, and so it's an honor to to be here. >> And I will say I hear a lot about >>
[08:01] >> Yeah. >> And I've learned a lot from Bruce over over a long period of time. There are There are certain people in this space There are lots of people who are visible publicly. There are also a group of people who are very visible privately or
[08:13] meaningful privately in terms of shaping how a lot of different people think and learn grow. I just put you in touch with someone today who's looking to learn more about the DC ecosystem, and you do a lot publicly, but there's a lot behind the scenes that you do that influence
[08:26] the ability of people to shape the policy discourse in positive ways. Um so, Alan, to your question, it makes me think about a moment that I had with my daughter this past January. We were walking in an icy place uh bringing
[08:38] pizza back to the place that we were staying, and we were all along a road, um and she was on the outside, and I said, "Why don't you take the inside cuz if a car comes and anything happens and it hits me,
[08:50] um I want it to hit me and not you." And and we were we were uh taking a pizza back from a from a restaurant, and she said, "That sounds great, Dad, but what about the pizza?" Hm. Um I think we're in a what about the pizza moment in AI policy where there have been larger
[09:03] providers who have been taking the hits, setting themselves up, I think, to take hits on the outside, and there's something important um that we need to preserve for smaller companies that I think is really at risk and is really challenging.
[09:16] The um the mantra that we've had since I started I I love it. I wish I could take credit for it, but people at our firm had developed it long before I started was focus on regulating harmful use of AI, not on regulating AI development.
[09:28] And so, the idea there was um everybody's got to comply with the law. The law covers a lot of harmful uses that we have in AI from unfair and deceptive trade practices to uh criminal conduct to violations of civil rights
[09:40] law. And if you enforce those things, you will protect people from harmful uses. But if you regulate development, so if you make it hard for people who enter a garage just to build the technology from
[09:52] day one, whether there are two people, whether there are there are 10 people, 20 people, whether they have any revenue in the door yet or zero, if you make that process harder, you will constrain the
[10:04] AI ecosystem and make it more difficult for us to have the kind of flourishing competitive market that we want to have. And I think unfortunately, we have seen just an enormous number of
[10:16] proposals and now increasingly enacted laws on the development side that make it harder for those two people in a garage to build new tools and get them off the ground. And I think that's a that's a challenging thing for us as investors, but I think more importantly, it's a
[10:29] challenging thing for the ecosystem, which we want to be diverse, and then for users, because users are beneficiaries of really competitive tech ecosystems. >> Well, to follow up on that thought, what is Bruce like what is driving this interest in trying
[10:42] to do more of the regulation and think about uh the kinds of things that Matt's talking about. I mean, everyone Washington, take Washington as example, this true in Brussels, this
[10:54] true all over. Everyone's talking about AI. Uh I will say not every policy maker is super deeply informed about how the technology works.
[11:06] But what is driving this I'm trying to be diplomatic. What is driving this interest? Like, you know, uh And what do you think is good What do you think is going to be moving in the near term or the in the medium term?
[11:19] >> Two very different questions. So let me first, what's driving the interest? Why have 488 million people apparently come to the Data Bricks conference? I mean, this is the hottest thing happening in in in capital markets, the
[11:31] hottest thing happening in technology, it's changing everything. Um you start with the People are really focused on this because this is a really big deal. So,
[11:43] sometimes there isn't necessarily the ulterior motives and we'll get into lots of those I suspect, but just start with policymakers should be focused on the biggest thing happening in every sector, in every country, uh in every part of
[11:56] our life, number one. Number two, many people within the uh AI sector are either busy saying how this is a bigger deal than fire. Like, well, why would
[12:08] you not expect to have regulators knock on your door if you're being if they're being told you've come up with something bigger than fire. And or there are also plenty of tech AI leaders and plenty of companies that seem to every week say,
[12:22] "We're all going to die. We're all going to be, you know, enslaved by robot overlords or something horrible is going to happen." Of course, if you're an elected official, it's a dereliction of duty if you don't listen
[12:34] to some of these, you know, Nobel Prize winners, really famous people, CEOs who are about to have IPOs in the trillions, and they're telling you, "We're really worried about this." You're doing your job if you're paying attention. But the
[12:47] last piece and the thing, you know, it's why it's so great that Andreessen Horowitz has a team, let alone a team led by people like Matt. Um you don't know. >> Yeah. >> And so, if any of us were in charge of,
[12:59] you know, you're an elected official, you're a representative, and you're asked to deal with something that's of high magnitude, that really informed people, some really informed people suggest could be really bad, and you don't know,
[13:11] what are you going to do? You're going to first do no harm. You're going to try to say, "Let's make sure this thing doesn't blow everybody up." And so, part of what we do, and certainly what you guys do, is you need
[13:23] to spend time with policymakers who don't have your background, don't have your experience, are listening to the CEOs on TV who maybe are, you know, marketing their products by talking about how dangerous and powerful they
[13:35] are, but you're going to have to just spend your time, and Alan, you know this is a is a former very senior government official, the amount of time when you were at the Commerce Department you had just go to the Hill and calmly and patiently
[13:50] for the 58th time explain why, here's what we're doing, here's why it's a good thing, don't believe what you're hearing. Um I don't know about what the roles of the folks in this room are, but I certainly hope and expect your companies
[14:03] you can either be heard or hope, and hope is not a strategy. >> Well, Matt, yeah. >> Yeah, can I come >> Yeah, and also just think about for the for folks in the room, and we have a collection of you know, there were over
[14:15] 31,000 people at this conference, right? At this summit. It's kind of incredible. People who are really building things. Like, what do you think respond to what Bruce said, but also like, what's important for the people who are builders to know and be thinking
[14:27] about in terms of what's going to happen? >> I so I think the real challenge for builders right now is that policy makers do not understand your equities. And unfortunately, I like I wish I had a better message on this. >> Yes, thanks. A ray of hope. Thank
[14:53] understand and incorporate those equities. One of my colleague Colin McCune wrote a piece a few months ago um about how policy makers will stand on a stage and say, we care about small businesses, we care about startups, and then they'll enact a set of policies
[15:05] that disadvantage startups who are going to compete with larger platforms. Previously you were alluding to policy makers understanding of AI, and I think policy makers lack of understanding of other technologies has really shaped in a negative way the policy discourse
[15:17] around around those technologies, but I'm not sure I don't I'm not sure the policy makers really understand AI, but I don't think that's the primary issue and gap that's going on right now. I think it's a failure um more from people who work in my ecosystem um in
[15:32] myself included to come up with the policy concepts that are responsive to the moment that when a policy maker looks you in the eye and says, "My constituents are scared about people losing their jobs." Our job is to come up with policy
[15:45] answers that are responsive to those to concerns, but don't the innovation that we think is going to bring a lot of value to people. Don't startups that are trying to compete with larger platforms. And we
[15:57] have tried really hard to suggest a set of policy alternatives that I think will push in that direction that can be responsive to the harms without undermining innovation and without making it hard for startups, but to be honest, I'm not I don't feel like
[16:09] they're breaking through right now. And as Bruce was alluding to, you hear from people including in industry this escalating rhetoric of how incredibly powerful and incredibly potentially damaging the technology will
[16:21] be, our the the ability of our message to hit home when we're when we say, "What you need to do is focusing focus on enforcing criminal law. You need to put enforcers in a position where they
[16:33] can really make sure when there are unfair and deceptive trade practices that they can take account of those harms and enforce against them." It's hard to break through when people are saying, "This is going to kill us all." >> The call is coming from inside the
[16:45] house. >> Yeah, mhm. >> Well, let me follow up on sort of like what are some good specific examples of things that policy makers could do that would help the cause. And I'm thinking about let me start with maybe asking you about open source, Matt, because you've written a ton. And Trisha and I always
[16:58] been a big and it's something that's by the way I should say near and dear to our heart at Databricks, right? I mean company founded you know, open data standards, open source software has been incredibly important to our development. Um we've
[17:10] open sourced a lot of our most even even this week we've open sourced some of our most important uh innovations. And open model uh open weight models are something that we've also uh done a lot
[17:22] of pioneering work on. I'm seeing some some pioneers here in the audience even. And so, um how do we How do you think about that as an example of an area a policy area where government could Should government
[17:34] be involved? Could government be involved? Is there a role? How maybe even just why is this so important? >> so it's vital to startups. It's vital to startups because not everyone can afford to build or pay for access to a proprietary model. And so, the research
[17:47] ecosystem and the startup ecosystem really depends on open source to flourish. You also, you haven't said this, but you also are a leader in thinking really deeply about open source from your time in the government. And I
[17:59] think the work that you did in the government is a really good example of a path that makes a lot of sense to follow. And I should say like this is not stories of bipartisanship or don't get headlines, but the the outline that you
[18:12] that you created, the sort of policy framework that you espoused, actually I think is largely consistent with how the Trump administration has opposed has approached open source. And I think that's a really positive thing. Right now, I think the data is really
[18:26] challenging in terms of American open source tools vis-a-vis Chinese open source tools. So, one of my colleagues, Martin Casado, has cited the statistic that of people you currently building on open source, 80% are building on Chinese open source.
[18:40] That is not, I think, where we want things to continue. That's not a status quo that we want to preserve. And I think arguably one of the reasons that we have gotten here is because regulatory uncertainty related to open source. There were people, including
[18:52] proprietary providers, who who suggested that open source might be less safe. I think that is probably untrue. And it's also certainly untrue, I think, that that or it's important to also emphasize
[19:04] that open source really has immense value for competition and innovation. And so, the framework that you outlined in the report that you helped to author when you were in government, I think takes the right approach, which is
[19:16] focusing on marginal risk related to open source and making sure that there's really to the extent that you're looking at risk that it's marginal and and that you're basing your assessment on evidence. And then I thought very intelligently the idea was to
[19:32] to not heavily regulate open source in the short run, but also to not like forswear ever regulating open source and to keep an eye on the development of the technology and when there's marginal risk that arises that needs to be taken into account, uh the government should
[19:44] do so. >>
[19:56] question. I maybe we can dig into some of the hot topics of the moment. Um I know that uh Mythos and Fable and Anthropic are very much on people's minds. So that's a uh I know it's a
[20:09] third-rail topic, but let's go there. Um For either of you, I mean I think the big question for the for folk maybe somebody can just sort of give a level set. I think probably everybody here has who's following the
[20:21] space has an under has some sense of what's been happening uh with the controversy and interest around Mythos, a very powerful model uh from Anthropic, and the government uh
[20:33] the federal government in the US recently issuing an export control order that made it, you know, basically impossible for uh Anthropic to continue to distribute and in fact on Friday uh
[20:46] evening they uh they stopped distribution and access to the Fable model. Does this in your mind represent a real sea change in the administration's approach? Is how big a
[20:58] deal is this? How big a disturbance in the force? Um should we be, you know, I guess one question is how how do you look at that uh development? >> I'm happy to start. Um,
[21:10] in some ways predictable. >> Yeah. >> You know, it's I mean, first Anthropic and the administration have been uh, fighting not least with the Department of uh, renamed Department of War, but um,
[21:22] they want to say, "If we sell it to you, we want to say what you can do or not can't do with the system." And the government the the Defense Department saying, "We buy F-16s from Lockheed, they don't tell us what missions we can
[21:34] fly on." So, we want access to these tools and that's that. So, that you start with there's a loss of love and a distrust, number one. Then number two, you know, Anthropic has spent a lot of time telling us all how insanely
[21:47] powerful Methuselahs and how much extraordinary damage can get done by it. So, now you know, you don't necessarily trust these guys. Now you're primed to think, "Oh, this is really scary stuff." They inform the administration of here's
[21:59] who we're going to give it to and then they added 50 more names and informed the administration after the fact of 50 more people they gave it to. I think when you look at the who's they've been sharing it with, very reasonable, but it didn't create a uh, uh, better
[22:12] communication or a greater trust environment. And so then when uh, you know, the Treasury Secretary hears that the that the fable version is he's told easily jailbroken
[22:25] and to uh, Matt's important point, policymakers don't understand this stuff. They're not going to they're going to listen to if a senior exec from Amazon calls, you assume they know what they're talking about. Amazon for that matter is an investor in and a partner of Anthropic,
[22:39] so it's not like they're you know, this isn't a dirty trick. It's a pretty reasonable reaction by the administration to say, "Let's hit pause because we don't want harm to happen and then it's very hard to put the genie
[22:51] back in the bottle and we don't trust these guys." Um, hopefully they'll figure it out. Uh, at the at the higher level what I worry about, I mean, software has eaten the world, not mine, because two reasons.
[23:03] You know, first, it's incredibly versatile and powerful, but second, the legal and regulatory model has been permissionless innovation. You come up with a kick-ass software, you have a five potentially five billion
[23:17] human totally addressable total addressable market immediately. Yeah, certain things can get can get dialed back, but part of the reason that's not going to continue is because of the technology is, you know, drones
[23:31] that can be set on auto kill, I mean, I think everybody in this room says, "Well, I hope hope somebody regulates that." You're similarly seeing if if there are tools that would give bad guys the ability to hack anything,
[23:45] government not knowing any better and not trusting necessarily the the party in question are going to hit pause. >> So, you're saying not a surprising reaction.
[23:57] Matt, I love your view of this. I mean, you know, the the other side in some ways is I can't think of a time in recent memory where the government has, you know, taken a piece of widely
[24:09] released software and said, "You can no longer distribute this." Is it It seems like a It seems like a real change for me. >> It reminds the the debate a little bit
[24:21] reminds me of a joke that a colleague told me. Um, person has a problem, person goes to the government, person now has two problems. And I I you know, Anthropic has like advocated for a policy agenda that included government having this power,
[24:33] and it's now being used against them. From our perspective, the policy set that is currently on the table, like the Overton window, the 50-yard line of where the policy agenda sits right now,
[24:45] has real costs. Um, I'm not a cybersecurity expert, so assessing exactly what the cybersecurity risk is is not my area of expertise, but the agenda that is being implemented to address that perceived risk has costs
[24:57] and has costs for startups. Um, Um and that's something that we're that we're really concerned about. I also wonder at times if if it is the most efficient way to go after the problem that we purport to care about. Um obviously our nation's
[25:11] critical infrastructure has vulnerabilities. It seems like that has the level of those vulnerabilities has become more apparent over the last couple of months and I'm not sure there's that much currently going on in the policy agenda that is really
[25:24] directly targeted at remediating those vulnerabilities and you can imagine a whole bunch of different policy concepts that would do that more directly like using AI tools that have been deployed to fix vulnerabilities in government's
[25:37] infrastructure and in private sector critical infrastructure. >> I want to turn just going to flag that we're going to turn to audience Q&A in a couple of minutes. So think about your questions. We'll have a mic going around.
[25:50] Um your chance to ask the ask the experts. Um for turning there maybe a sort of a question about uh Bruce you you you had a you talked about Dickens. So like you
[26:02] know I I in some ways one could argue this is a tale of two cities right now in the public discourse and conversation about AI policy, right? On the one hand I think we sit in rooms where people
[26:14] talk about permissionless innovation where industry is very I think strongly aligned in the view that and many people across discourse are really strongly aligned in the view that um we need these these tools are going to be so
[26:26] beneficial so powerful. We really need to try and get them into people's hands as quickly as we can and do everything we can to develop them uh and and and support their development and we are investing trillions literally
[26:38] trillions of dollars in infrastructure uh across our economy. The other piece of the conversation is very different. It's the the fear uh the worries about the implications on
[26:50] the labor market on the um uh you know, on privacy, on security, um uh on our civic discourse. How do you see these this tale of two
[27:02] cities resolving? Right? Like, what where you know, uh how does this play out? And we're seeing it around the world in a geo with a geopolitical overlay, too. Right? This is not just in the US, it's in Europe.
[27:15] China's plays a big role, um the competition with China from the West. How do you see these tensions resolving over time? Nice easy topic. Um but the tale of two cities, how does this
[27:27] conversation get bridged? >> Well, first, two things can be true. It can be uh extraordinary technology that will uh change lives in the best of ways, like disease discovery. And it can be extraordinarily powerful
[27:39] technology that creates greater risks than we've seen. I think both things are true and and will remain true. You know, how historically these sorts of things Do you ever have your moderator leave the room at the times? >> Uh sorry, I should have said I'm getting
[27:51] my the mic for questions, so >> If you ask questions from there, it's not audience questions. It's
[28:03] we're going to be in a uh the era of straight-up permissionless innovation is more over. There are going to be certain uses of the technology that are going to be prohibited. I'm hoping, to your point, um development is
[28:16] less constricted, though it may depend on development of what? And if you're explicitly developing, you know, we want to build something that will combine AI and CRISPR to create bioweapons that there are no cures for, I kind of hope
[28:29] the development of that gets blocked at some point, too. My take is, you know, I I Jonathan, we were talking about this dinner last night. I mean, there's the 20% oh, it's all horrible, and the 20% um we we never work again, and everything's
[28:42] awesome. And I'm in that firm, you know, middle 60% of we're going to genuinely generally model through. We're going to figure out regulations that that uh protect people from the worst of the
[28:55] stuff. But we're if the industry can be smart and work with government to make sure that we don't have both high risk and extraordinarily narrow winner circles.
[29:08] That if the benefits are broadly both understood and broadly enjoyed, I think uh this is ultimately a massive societal win and an industry win. >> Uh I think if it's if it's a tale of two cities, one of them is very sparsely
[29:21] populated and seems to be on the decline. Um I think permissionless innovation, unfortunately in the current environment, won't land as a message. I think it is essentially like dead on arrival. Um and the challenge in the current movement is coming up with the
[29:34] policy concepts that are responsive, I think, to the the tail in the other city, um but that don't have some of the crippling effects that I think we're all concerned about. >> Please join me in thanking our panelists today.
[29:46] >>

Learn more about the Databricks Data and AI platform.

The information provided herein is for general informational purposes only and may not reflect the most current product capabilities or configurations.