AI Horizons Summit with Gov. Josh Shapiro in Bakery Square in Pittsburgh.
00:00 - (upbeat music)
00:02 - - Please welcome Governor Josh Shapiro.
00:06 - (audience applauds) - Thank you.
00:08 - (audience applauds)
00:13 - - Thank you everyone, thank you everyone.
00:16 - Thank you for being here and I wanna begin
00:20 - with a special thanks to Joanna for all of the good work.
00:26 - (audience applauds)
00:30 - All the good work you've done again and again
00:32 - through the AI strike team to convene industry leaders
00:37 - for important conversations like the ones we are having over these couple of days.
00:42 - You know, I have joined you at this conference as governor
00:46 - each year to talk about the future of artificial intelligence
00:52 - because it has been clear to me for years that the technology has the potential
00:57 - to reshape our society in really profound ways.
01:03 - Just last year, here on this stage,
01:06 - I talked about the promise and the possibility of AI
01:11 - and what it means for the Commonwealth of Pennsylvania.
01:15 - I talked about how AI can create efficiencies,
01:18 - better health outcomes, assist workers in their daily tasks
01:23 - if it is developed in responsible and ethical ways.
01:29 - I still believe that that opportunity exists.
01:33 - But as the events of the past several weeks
01:36 - have made even clear, we cannot ignore the
01:40 - warning signs nor downplay the very real risks.
01:45 - Let's talk about what has changed over the course of the last year since we last met.
01:51 - For one, the capabilities of this technology
01:54 - will they have advanced significantly.
01:57 - And second, the technology is now starting
02:01 - to improve itself, oftentimes escaping the controls of its human developers.
02:08 - More people are using artificial intelligence than ever before.
02:14 - Earlier today, I was with a plumber who told me
02:18 - that he uses AI to improve his small business.
02:22 - In the next breath, he told me how scared he is
02:26 - about the future direction of that technology.
02:29 - You see, that plumber understands instinctively
02:33 - what is being talked about in these halls today
02:36 - and what is being talked about across America today.
02:39 - The pace of development of these frontier models is getting exponentially faster.
02:45 - And now, the leading companies developing these tools say that
02:51 - they can't guarantee that the public will be safe from them.
02:56 - What we've learned about AI agents forming their
02:59 - own clicks developing their own language and trying
03:03 - to mislead their human developers while hacking
03:07 - into other companies' networks on their own,
03:09 - most notably in the hugging face earlier this year.
03:14 - And just yesterday, open AI reported six additional instances
03:20 - just in the last six months of these agents acting outside of their control.
03:27 - We've heard warnings about how these tools could
03:29 - be used by bad actors to build biological weapons
03:33 - or to surveil or suppress political opponents.
03:38 - All the while, leaders in our federal government
03:41 - refuse to put in place any guardrails to ensure this technology is used responsibly,
03:49 - whether out of willful ignorance or whether out
03:52 - of personal greed.
03:55 - I want to be clear, I am not a doomer.
03:59 - I don't think it is inevitable that AI
04:02 - is going to take over the world where
04:04 - that someone will misuse it to kill us or to do harm.
04:10 - But I do have real concerns, as do many Pennsylvanians and many Americans.
04:17 - And I don't think that the very real concerns of our neighbors
04:21 - that they overwhelmingly express about this
04:24 - technology is a hoax, as our president has called it.
04:29 - The leading companies in this space and some who
04:33 - are here with us today have openly talked about what
04:37 - could happen if this technology falls into the wrong hands
04:41 - and how more and more their own hands can't fully guide it and shape it.
04:48 - The fear and confusion that every day
04:51 - Americans have is understandable, and it needs
04:54 - to be addressed comprehensively with a focus on responsible,
05:00 - safe, and ethical uses of this extraordinary technology.
05:05 - Here in Pennsylvania, well, I believe
05:07 - we are uniquely situated to be a leader in this conversation.
05:12 - We've got world-class institutions,
05:15 - like Carnegie Mellon, Pitt, the University of Pennsylvania
05:19 - and Penn State, where some of the smartest minds
05:22 - have been thinking about these important issues for years.
05:25 - We've got business leaders, organized labor
05:28 - and community leaders, all getting together around the table
05:31 - that convenings like this to have hard conversations
05:35 - and, most importantly, to collaborate with one another.
05:39 - And I want you to know you have a governor that
05:42 - is ready to engage on these very tough issues.
05:46 - I know there are many companies and researchers,
05:49 - including many here in this room, who
05:51 - are working on AI technology that has the potential
05:55 - to do so much good in the world.
05:57 - Some are already improving people's lives, companies
06:02 - and researchers who are deploying this technology
06:05 - to solve real problems in life sciences,
06:08 - in agriculture, in infrastructure, and in many other fields.
06:13 - Take my friend, Jake Lucerarian, who you heard from
06:17 - a few moments ago.
06:18 - He's one example.
06:20 - He found a gecko robotics here in Western
06:22 - Pennsylvania, and he's using AI embedded in his robots
06:27 - to inspect critical infrastructure, like power plants,
06:31 - to ensure that they are safe and operating efficiently.
06:36 - He is helping keep our community safe and ensuring
06:39 - that when we flip on a switch in our house, the lights actually turn on.
06:44 - Gecko is a Pennsylvania success story.
06:48 - Jake built that company with 10,000 bucks in seed funding
06:53 - from our Commonwealth, and he's turned it into more than a billion dollar enterprise,
06:58 - based right here in the steel city, with hundreds of employees around the world.
07:03 - Here's one other example.
07:06 - The human engineering research labs at the University of Pittsburgh.
07:10 - They're using AI to develop a wheelchair with a robotic arm
07:14 - that can do everything from open doors to carry a beverage,
07:19 - giving folks with limited mobility, real independence.
07:23 - And let me give you just one more example of many.
07:27 - A Philadelphia-based company called Burrow,
07:29 - which is using AI to build autonomous robots that can help our farm workers
07:34 - haul and produce more, haul produce more quickly safely
07:40 - and efficiently, getting more food on Pennsylvania's plates much faster.
07:45 - Those are just three examples of companies and
07:49 - researchers who are responsibly and ethically utilizing
07:54 - the promise of this technology to create opportunities, not to limit them.
08:00 - Three organizations that are using their big brainpower
08:04 - and their vision to deliver tangible benefits to real people.
08:09 - We need to ensure that this technology benefits all
08:13 - Pennsylvanians and doesn't just enrich the people who build it.
08:19 - Unfortunately, too many companies aren't
08:22 - using these technologies ethically or responsibly,
08:27 - either because of purposeful actions or because
08:31 - they've lost control of what they started to build.
08:37 - They're focused on winning the race,
08:39 - developing the next model or growing at scale above all else,
08:44 - and not always paying attention or caring about the negative consequences.
08:50 - Let me give you an example of a company we've encountered here in Pennsylvania
08:55 - that's using this technology in a way that harms others.
09:00 - The use of AI companion bots is growing rapidly.
09:04 - In fact, about 30% of teens across the country
09:07 - report using an AI companion bot each day.
09:11 - But this space is entirely unregulated.
09:15 - We've read stories about these chatbots telling
09:17 - kids to hurt others or even to hurt themselves.
09:22 - And so we investigated through our state
09:25 - department of state who oversee this type of work.
09:30 - Within just a few minutes, our investigators,
09:33 - using their official state email addresses, were connected with a chat bot who
09:38 - went by the name of Emily, who claimed to be an expert in
09:43 - psychology and psychiatry and quickly dispensed medical advice.
09:49 - The chat bot even produced a bogus medical license
09:53 - number in Pennsylvania presenting itself as a licensed medical provider.
09:58 - That is against state law.
10:01 - And my administration sued character AI to stop it.
10:06 - You know my track record as your attorney general.
10:09 - I will go after anyone who harms our children.
10:13 - But we can't just do this retroactively
10:16 - and wait for people to get hurt or wait for these companies to break the law.
10:21 - We need to put proactive measures in place and establish responsible regulations
10:28 - to prevent this kind of thing from happening in the first place.
10:33 - In fact, that is what we've been doing
10:36 - since the first few months since I took office as your governor.
10:40 - When I signed an executive order establishing key principles
10:45 - and setting up real guardrails governing how Pennsylvania would use generative AI.
10:51 - Those principles, respecting privacy, ensuring safety
10:56 - and security, promoting fairness, increasing
11:00 - accuracy, and empowering our employees.
11:03 - That has continued to guide every decision
11:06 - we've made about AI use in state government.
11:10 - At the same time, we've established partnerships
11:13 - with CMU and Penn to tap into their networks of experts
11:18 - and researchers to develop proactive forward-looking policy.
11:24 - We launched an AI pilot in state government
11:26 - that was grounded in those values and shaped by sound policy.
11:32 - We engaged our labor leaders and Commonwealth employees
11:36 - to get their buy-in and we ensured everyone using this technology had proper training.
11:42 - And through that, we sped up government and we did so in a responsible ethical way.
11:48 - As a result, our state employees found the tools exceptionally useful.
11:53 - We eliminated backlogs.
11:55 - We sped up permitting and our employees
11:58 - reported saving 95 minutes a day using them.
12:02 - Pennsylvania, well, we have been a model for safe and effective AI deployment
12:08 - to our Commonwealth employees from the beginning.
12:11 - But I'm mindful that we are just one state.
12:14 - And despite our commitment to collaboration and the leadership of folks in this room
12:20 - and throughout the region, we can't address this alone.
12:24 - It's going to require national, in fact,
12:28 - international, leadership and collaboration.
12:32 - I've called on the federal government to step up and regulate the AI industry.
12:37 - Just like they do for nearly every other
12:40 - industry that significantly impacts our lives.
12:44 - Think about it.
12:45 - We don't let pharmaceutical companies develop
12:47 - drugs and put them right out in the market.
12:50 - No, they must first go through the FDA to prove that they are safe and effective.
12:56 - And we don't let aircraft manufacturers make an airplane
12:59 - and then just put commercial passengers on it right off the assembly line.
13:04 - No, their, the FAA said strict standards that must be met before that first flight.
13:11 - I believe AI should not be treated any differently.
13:15 - Particularly because it has the power to
13:18 - impact people's lives in such dramatic ways.
13:22 - We need the federal government to establish strict rules
13:25 - of the road and then enforce those rules to ensure a level playing field.
13:31 - Rules that require independent third-party oversight
13:36 - of frontier models, increase transparency into the development
13:42 - of these models and prioritize safety above all else.
13:47 - The future of a technology this powerful should be
13:51 - shaped by the American people and their representatives.
13:55 - And while information from industry leaders is helpful
14:00 - and very welcome, they should not exclusively
14:03 - guide the regulatory process of this nation.
14:07 - The federal government needs to work with leading researchers and experts
14:12 - as we've done here in Pennsylvania and with the American people, not against them
14:19 - to develop strict guardrails on this technology,
14:21 - its development and its deployment in the real world.
14:26 - Look, I recognize that that is hard work.
14:29 - But we elect our leaders to take on the hardest problems,
14:33 - not to give up at the first sight of difficulty.
14:36 - Speaker Mike Johnson and too many in Congress
14:40 - will they are content to do absolutely nothing
14:43 - instead blaming others for their lack of leadership
14:46 - or saying that they don't have the expertise to act.
14:50 - Think of this, just days after the leaders
14:55 - of all of these top technology companies in the world issued apocalyptic warnings.
15:02 - Mike Johnson adjourned the House of Representatives
15:05 - last night, weeks earlier than originally scheduled.
15:09 - So members could go home and campaign
15:12 - instead of staying in Washington and doing their jobs.
15:15 - He has abdicated his responsibility to lead at a
15:19 - time when we cannot afford weakness and retreat.
15:24 - By the time Congress comes back into session after the election,
15:29 - we could be dealing with a whole new generation
15:32 - of generative AI and with more scary examples
15:36 - of AI agents going rogue.
15:40 - But there's something in between accelerating
15:43 - and ignoring the road ahead and slamming on the brakes and refusing to move forward.
15:51 - It's a false choice to say that our only two options
15:55 - in America are to do what Donald Trump is saying
15:58 - that we should ignore the risks and go full steam ahead
16:02 - or on the other hand to completely stop work on AI development indefinitely.
16:08 - There is a better, more thoughtful way.
16:11 - And that is what we have to collectively work toward.
16:16 - Strong federal action is the first step.
16:19 - But this issue and this technology cannot be contained within our national borders.
16:25 - As I've said before, and believe to this day, we need to beat China on AI.
16:32 - I believe that because I trust American workers,
16:36 - democratic ideals to develop these technologies more
16:40 - responsibly than the developers in communist China.
16:44 - But the idea that we have to choose between
16:48 - protecting our people from the risks posed by AI
16:53 - and winning the race against China, well, that is just wrong.
16:57 - That is another false choice.
17:01 - Because the way we win, both on AI and to protect the American people,
17:07 - is to a process of building international consensus
17:12 - around this technology and its use, to
17:15 - establish common protections for our citizens
17:18 - and a level playing field for our companies at home and abroad.
17:24 - We should work with our allies and our
17:27 - adversaries to put these protections in place.
17:32 - That might mean placing some limits on technology
17:36 - sharing and slowing down frontier AI development
17:40 - until we can be sure that these systems are safe.
17:45 - And while China may complain that we are, quote, fear-mongering, both our countries
17:51 - have a mutual interest in getting this right and ensuring
17:55 - that we don't lose control of this powerful technology.
17:59 - Unfortunately, when you isolate yourself
18:03 - as the Trump administration has done to our country over the past several years,
18:07 - it becomes harder to build that necessary international consensus.
18:12 - And while the President talks a big game about beating China on AI development,
18:17 - remember it was his administration that cut a deal
18:22 - to sell the most advanced chips to Chinese companies
18:26 - giving them a leg up over American entrepreneurs.
18:31 - That was a mistake.
18:33 - Real leadership means keeping America ahead of China in AI innovation.
18:39 - It means pushing Beijing on practical guardrails
18:42 - to reduce risks to the American people and putting
18:46 - effective safety measures in place right here at home.
18:50 - It's time for the administration in Washington to stop alienating our allies.
18:56 - Starting wars that destabilize our world
18:59 - and damaging our credibility in foreign capitals.
19:04 - Instead, they should start leading the conversation
19:07 - on how we can mobilize countries around the
19:10 - world to join us in responsibly regulating AI.
19:15 - That should start next week.
19:17 - When President Xi visits the United States, this should be among the top issues
19:22 - on President Trump's agenda for their visit.
19:26 - I for one hope that President Trump
19:28 - will leave the bluster in the showmanship at the door and engage proactively,
19:34 - productively and responsibly on this issue that is so critical to our nation's future.
19:42 - For my part, I'll be leading the conversation here
19:45 - in Pennsylvania for safe and responsible uses of AI.
19:50 - And I'm here today to respectfully ask everyone
19:53 - in this room to be a part of that effort.
19:58 - Let's build on the good work that is
20:01 - happening here at this AI Horizons conference.
20:04 - This is the foundation upon which we can do this hard work.
20:09 - I know that some of you are thinking
20:11 - that conversations about regulation
20:14 - will slow us down or undermine our competitive
20:17 - spirit or force us to have to make compromises.
20:21 - But that is not the case.
20:25 - The companies that will really succeed in the years to come,
20:29 - they won't be the ones that bet on the Wild Wild West.
20:33 - They'll be the ones that understand that swift progress
20:37 - grounded in a commitment to responsible
20:41 - ethical development, that that is the answer.
20:45 - I appreciate the concerns that have been
20:47 - articulated by the CEOs of these leading AI labs,
20:53 - including several companies who are represented here today
20:57 - and participating in this conference.
21:00 - But we should be clear.
21:01 - Some of those same companies who have been calling
21:05 - for regulation in public have been funding the groups
21:09 - in private who have been fighting against it.
21:12 - They don't get to claim the moral high ground
21:14 - by saying one thing publicly and doing another thing privately.
21:19 - Because people are pretty smart.
21:22 - They can see right through that.
21:24 - They can tell when a CEO is more concerned about their IPO
21:28 - than the impact their product is happening on
21:31 - our kids, our families, and our communities.
21:36 - To address the significant concerns that people have with the AI industry,
21:43 - the folks here in this room and folks in boardrooms
21:47 - in Silicon Valley will they need to take action too.
21:50 - I believe you have a social responsibility and you need to walk the walk.
21:56 - You need to show how your technology is helpful to real people.
22:01 - The way Jake is over at Gecko or Dr. Cooper is
22:05 - at Hurl where the team at Borough have done.
22:08 - Participate in the regulatory process.
22:12 - Help frame it, but don't fight it.
22:15 - Invest in the advancements that make people's lives better.
22:19 - Operate transparently and make safety your first priority.
22:24 - Focus on creating outcomes.
22:27 - They're gonna actually solve problems.
22:29 - They're gonna keep people healthy.
22:32 - Make inefficient systems work better and otherwise improve people's lives.
22:39 - And do that in an ethical, responsible way
22:42 - that protects jobs and doesn't eliminate it.
22:46 - That should be our primary focus, not who crosses the tape first.
22:53 - This is not a moment when we can walk away
22:56 - from hard conversations or from emerging technologies.
23:01 - We can't kick the can down the road anymore.
23:04 - And I chose to come back here today at Joanna's invitation
23:08 - because I believe this group has a lot to offer the national conversation.
23:14 - In this room, we have experts in technology and
23:18 - business and academia in labor and in government.
23:22 - We've shown that we can come together as a community
23:25 - and we've shown time and time again that we can work together responsibly.
23:30 - And while the federal government quite
23:31 - literally walked away from the table this week,
23:35 - we need to step up and be a model for the rest of the nation
23:40 - and force them to come back and do their jobs.
23:45 - And listen, I'm not asking you to do charitable work.
23:50 - I know this isn't a convening of nonprofits.
23:53 - Everyone in this room can be successful with this technology.
23:59 - But the people who will lead this future field,
24:03 - the companies that earn the people's trust in Pennsylvania
24:08 - will those will be the ones who do it the right way.
24:11 - Those will be the ones that ultimately profit.
24:14 - Those who abide by responsible guardrails who prioritize
24:18 - safety and who use this incredibly powerful
24:22 - tool to make a difference in people's lives.
24:26 - If you go back and look at the commercial history of Pittsburgh,
24:32 - it's filled with companies that have solved problems
24:36 - and stake their reputations on doing it the right way.
24:41 - We gather here today to talk about the frontier of innovation,
24:46 - but 157 years ago, another technological pioneer
24:52 - was founded right here in this city, the Westinghouse Air Break Company.
24:59 - George Westinghouse, one of the greatest inventors of his time,
25:04 - built his company to solve a problem that he witnessed with his own eyes.
25:12 - Westinghouse literally saw two runaway trains heading right toward one another.
25:18 - Their engineers tried to stop them, but they couldn't break in time.
25:24 - Mr. Westinghouse saw that crash.
25:26 - He reacted and he invented new breaks to stop trains that would
25:30 - ultimately be used on locomotives all across this country.
25:36 - He built an incredibly successful company
25:40 - and he did so by treating workers with dignity and respect.
25:45 - At a time when his fellow industrialists
25:47 - were busting strikes and undermining worker rights,
25:51 - Westinghouse gave his workers more benefits and more protections and ultimately solved
25:57 - a critically important problem in this nation.
26:00 - He judged himself, not on how much money he made
26:04 - or how many inventions that he patented, but whether his work, quote, "contributed,"
26:10 - something to the welfare and happiness of humanity.
26:15 - Think of his words today.
26:17 - It is in that spirit that we must address these challenging questions of our time.
26:23 - We must work together, treat one another
26:27 - with respect like we always have here in Western Pennsylvania.
26:31 - It's part of our DNA to build trust to establish responsible guardrails,
26:37 - to ensure the technology is safe and that it benefits us all.
26:43 - Pittsburgh and Pennsylvania can lead the way.
26:47 - We can be a model for the rest of the nation.
26:51 - People are counting on us and I believe now is our time.
26:56 - Thank you all very much.
26:58 - I wish you the best with this work and I'll be with you every step of the way.
27:02 - Thank you.
27:03 - (audience applauding) Thank you.
27:06 - (audience applauding) Thank you.
27:09 - (upbeat music)
27:21 - (upbeat music)