Tonight, I’d like to talk about trust, democracy and media, and the evolving role of digital platforms and First Amendment rights and values. I’d also like to tell you a little about what we in philanthropy are doing about it.

Thanks to the Salzburg Global Seminar for inviting me to speak tonight. In a world of new rules and lightning-fast communication, your role as a haven for thoughtful exploration of complex issues has never been more important.

Let me begin by focusing on the technology companies that play such a dominant role in our media landscape. A few weeks ago, representatives from Facebook, Google and Twitter came to town to testify before Congress. In sober hearings, our representatives peppered them with questions, largely aimed at understanding Russian social media activities during the 2016 election.

That inquiry is important, but let’s look beyond the narrow scope of those hearings and explore a broader conceptual issue, a massive and thorny topic: the role and responsibility of technology companies that began as platforms and transformed, I believe, into publishers. These are two very different things, with different roles in society. Are they merely platforms and tech companies, or are they publishers with social and legal responsibility for what they publish? That is a central question at the heart of how to use internet for democracy, and it involves technology, evolving attitudes toward First Amendment values, and key questions challenging big tech’s business models.

Throughout history, humans have grappled with how to identify truth, control information, and empower people with knowledge. The Greeks struggled to balance common identity and purpose with free and democratic expression. We can each point to a different period in history when technology complicated and charged that quest, sometimes using information for good, and sometimes for evil.

As we think about this, let’s take some comfort from Gutenberg. Before Johannes Gutenberg mechanized the printing press, there was order. Books were rare, distributed from the few to the few, and they usually came with a cardinal’s imprimatur asserting truth. After Gutenberg, any Tom, Dick or Martin Luther could print and distribute whatever he wanted. Information flowed from the few to the many, and soon from the many to the many. So many, in fact, that information and opinion became hard to control, unreliable and unruly. It took a hundred years and evolving experience with technology and its governance for people to relearn to trust information.

But trust they did, eventually. And we are at a similar place.

In a room of people who know Bob Schieffer, I should note he makes this case in his new book, “Overload.” And if you want to go deeper, see University of Michigan professor Elizabeth Eisenstein’s “The Printing Press as an Agent of Change.”

At the beginning of our own republic, the reach of media was local and largely verifiable. The public learned to trust information because they could see for themselves whether it was in fact true. The circulation of leaflets and newspapers extended to roughly the area of electoral districts. The Founding Fathers publicly debated the core ideals of our republic through accessible argument. In doing so, they formalized the role of the press as the staging ground for the middle, an arena of words where common ground is the common prize, where left and right could come together in compromise.

And it remained that way, with newspapers, pamphlets and later, radio and television informing communities, though of course not without the speedbumps of a sometimes partisan, sometimes warmongering, sometimes bigoted press, reflecting their owners and the times. But by and large, the U.S. grew up with local papers that established a direct relationship between themselves and their communities. That relationship held through most of the 20th century, until the phenomenal rise of internet.

This discussion would have been difficult to imagine a few decades ago, before the first electronic message traveled between two computers; or 15 years ago, before Facebook; or a dozen years ago, before the first tweet. But, conceptually, the issue is not new: technology has upended society many times before. And internet represents the most fundamental change to media and society since Gutenberg. It is both the greatest democratizing tool in history and democracy’s greatest challenge. It gives us all voice and potential influence beyond previous imagining.

And, at the same time, with the country dividing itself, in real life and online, into the more homogeneous communities described in the “The Big Sort,” internet has facilitated the creation not just of the filter bubbles but of the protective shields that allowed us to block dissenting or differing views.

The success of any news operation used to be measured by its ability to effectively and reliably inform society. But the business model that sustained newspapers for more than a century is now broken. Gathering and disseminating accurate information is expensive and revenue is short. We now have a simultaneous torrent of individual and small information efforts, and a potentially dangerous concentration of power in the hands of a handful of private companies with seemingly boundless potential for reach. “Media” now means digital and cable, cool mediums that require hot performance.

Americans’ trust in institutions, and in each other, is at historic lows. According to Pew, only 20 percent of Americans trust their government. The same low percentage has “a lot” of trust in the national news media. I agree with Nina Jankowicz of the Woodrow Wilson Center, who recently wrote in the New York Times that:

“It’s impossible to say definitively what causes this mistrust, but its growth has coincided with the rise of both the adrenaline-driven internet news cycle and the dying of local journalism over the past two decades. Without news that connects people to their town councils or county fairs, or stories that analyze how federal policies affect local businesses, people are left with news about big banks in New York and dirty politics in Washington.”

Said another way, there simply are fewer and fewer institutions unifying community by feeding news and information to the middle and setting the agenda for civil discourse.

Social media has catalyzed the fragmentation of what was once a somewhat unified public sphere; it has fractured and privatized the town square. The shattering of communal baselines was becoming a problem before the social media boom — but the divides among us have only grown starker, as we find less and less common ground and rely more and more on opinion presented as news. Earlier this year, Knight Foundation partnered with the Aspen Institute to form a Commission on Trust, Media and Democracy to consider these questions. Led by New York Public Library president Tony Marx and former Tennessee state legislator Jamie Woodson, the Commission will consider the fundamental issues of trust and recommend solutions to restore it.

What we know, or think we know, which leads to what and who we trust, and who we deem trustworthy, is increasingly determined by five behemoths: Facebook, Google, Amazon, Apple and Microsoft. As examples of their dominance, consider that Facebook and Google capture more than 75 percent of digital advertising revenue, and make up 40 percent of America’s digital content consumption. In another time, we might have looked at such dominance with the same jaundiced eye that Presidents McKinley and Teddy Roosevelt used to gaze on Rockefeller’s Standard Oil or Carnegie’s U.S. Steel. So far, we haven’t. But that could change.

Silicon Valley has shied away from accepting a publisher’s basic responsibility for the authenticity of their content. They have virtually limitless opportunity to define and form community beyond geographic boundaries. But they disavow responsibility for authenticity — for the truth or falsity of the material — as a basic tenet of their business model.

A publisher’s success is premised on consistently delivering reliable news and information, while today’s platforms premise success on basic access tailored to personal preferences for any proposition, person or group. Those are new rules indeed! And, I think, they may be unsustainable.

Jack Knight’s notion of a well-informed citizenry eager to tackle questions within a common factual framework is challenged today. With the disaggregation of news sources and the rise of technology companies as leading publishers, Americans have lost the bedrock of democracy: a shared baseline of facts.

Today, people already view tech companies as media companies. Pew research shows that a majority of American adults get their news on social media. Facebook is the top source of political news for the millennial generation. In all age groups, the percentage of social media users who rely on those platforms for news is increasing.

This dominance is no accident.

The amount of information about ourselves, our habits and our preferences that we share with these tech companies in exchange for convenience is stunning. In addition, tech companies already produce and will produce more and more content: Think of YouTube News and Facebook paying to create video content to share on their live platform, or Amazon Studios. These companies may shun the media label, but they proactively pursue media revenue streams.

Like the trusts of decades past, the tech giants are horizontally integrated, saturating and dominating the market for information, and sometimes vertically integrated, exerting influence and control over news at every stage, from generation and production through distribution.

Given their success, you can’t blame media platform companies for not wanting to change. They never intended to shoulder responsibility for reporting the news. But sometimes life’s unfair and takes an unexpected turn.

I think they’ll change because they’ll have to. Because the world and their role in it have changed from the time when they were fledgling startups to a time when, for example, Facebook and Google have become more influential purveyors of information than the New York Times or the Washington Post.

I think they’ll have to because it’s bad for business if people lose trust in what they read on Facebook. People don’t say, “I read so-and-so’s article from such-and-such publication.” They say, “I read on Facebook that … .” And if it turns out that they can’t trust what they read on their Facebook newsfeed, that’s bad for business because it may mean a customer who’ll look elsewhere for information they can trust. I believe it’s in Facebook’s self-interest to deal with this trend.

How fast might there be a consumer revolt? No one can say. But no one can also deny that it could happen quickly. To borrow from Malcolm Gladwell, I can imagine a tipping point when our society says “enough” and consumers change the game. History is replete with these examples. I think we’re living in such a moment now in the U.S. around sexual harassment and abuse. After decades of silence and looking the other way, those stories are like kindling, and the Harvey Weinstein disclosures were the spark that set the fire of change. I truly believe we’re witnessing fundamental change in attitudes and practices.

The point is, people have always had the power of collective will, and with the speed of communication now, it is potentially lethally quick. In consumer markets, when customers feel they’re not well served by existing services, they’ll find others, just as they found Facebook instead of MySpace or Google instead of AltaVista. Yogi Berra, as usual was right: “If the fans don’t want to come to the ballpark, nobody can stop them.”

Another reason the platform companies will change is because technology will enable them to assume more effective control of their content, which presents all sorts of other questions about machine learning and decision-making. The technology isn’t there — yet — to exercise human judgments, but it is already there in terms of processing data on machine scale, outstripping any human capacity. I think it’s just a matter of time.

You may see a change in the big tech corporations forced by government for two reasons. First, the power of these media companies in society is classically the kind of monopolistic or near-monopolistic power government traditionally protects us from. Government could ultimately trust-bust. We’re not there yet, but the word antitrust is much more in use now than I was hearing even just a few months ago.

Secondly, government might force change if the public got sufficiently riled up about the lack of reliability — the lack of authenticity — on internet. If internet media corporations declined to accept even modest responsibility for what appears on their sites, governments may gladly step in to correct and possibly control, and I don’t just mean in China, where censors effectively limit what can be seen by a huge portion of the world’s online population.

In the meantime, who is doing what to respond to these challenges?

I give Facebook and Google, in particular, credit for now taking this matter seriously. What they’ll ultimately do and how they’ll do it is unclear to me, but I believe they understand the world has changed.

Others, like The Trust Project, supported by Craig Newmark of Craigslist, Google and Knight Foundation at Santa Clara University, are working with newsrooms and technology companies to help the public and algorithms differentiate between news content and fakery. A proposed new project, NewsGuard, spearheaded by Gordon Crovitz (formerly of the Wall Street Journal) and journalist-entrepreneur Steve Brill seeks to make this effort even simpler for the public — and technology companies — by ranking news sites based on their authenticity track record with clearly visible green, yellow and red signals to alert users. These efforts, and others like them, are not perfect, but they should act as a not-so-subtle reminder to the user to consider authenticity and ultimately, demand it.

At Knight, we believe artificial intelligence will play an increasingly central role in news. With Pierre Omidyar, who co-founded eBay, and Reid Hoffman, who co-founded LinkedIn, Knight Foundation has started a fund to work with MIT’s Media Lab and Harvard Law School’s Berkman Center to explore the ethics and governance of artificial intelligence. When those tech solutions I just suggested are developed, how fast can they become dark implements in the hands of Big Brother? In a New York heartbeat. How obscure can tech companies be? Darker than Darth Vader. And they speak a language most of us, including members of Congress, don’t fully understand.

So it’s critical that organizations like ours, and others like the Hastings Center or MacArthur Foundation be encouraged to do more. So, too, should Washington think tanks and research universities. Leaving the ethics and governance of how we know what we know to corporations whose primary purpose is commercial gain, or to whoever is in power in government, is not just bad policy for the short term, it’s bad for democratic society in the long term.

In the meanwhile, what are traditional media organizations to do? Again, I look for green shoots and find them in organizations like the Washington Post, The Texas Tribune and ProPublica. These organizations have different business models but share a common commitment to verification journalism, and to the use of technology to find and reach their audience. They believe, as my friend Marty Baron likes to say, that we have to focus on doing the work of getting the story right, because that’s where credibility will be proven and earned.

Knight, along with the Lenfest Institute in Philadelphia, has been working with dozens of newspapers and some television stations to help them transition to digital. This shift requires culture change, a shift from news by appointment to news any time it’s verified; a shift from “I write/you read” to letting readers enjoy the benefits of a platform. As much as I think digital platforms have to accept publisher responsibilities, I think traditional publishers need to adopt the techniques of digital platforms, and their relationship to readers and users, in order to compete. Digital technology is changing us and the truth is, as a friend told me recently, we’re in kindergarten, trying to get ready for graduate school.

Since we are in the Newseum, let me end with some observations about the changing understanding of what free speech and free press mean to young Americans. Although the ferocity, reach and frequency of today’s political attacks have ratcheted up the level of intensity, I’ve talked with too many politicians at too many levels, local and federal, to believe their view of media is fundamentally new. What I think is new is a changing generational view of what free speech means.

In early 2016, Knight commissioned a Gallup survey of attitudes among college students toward First Amendment freedoms. The results suggest that there is a fundamental, generational shift in our understanding of these basic rights among a broad and deep sample of young people training to become our nation’s leaders.

The Gallup survey showed that about three-fourths of all college students believe in free speech, but about two-thirds believe in “safe spaces.” Do the math and scratch your head. And among a representative sample of African-American college students, some 60 percent did not believe their right of assembly was secure.

My own reading is that this younger generation of leaders values inclusion and freedom from psychological harm in the same way that previous generations valued freedom from physical harm. In the traditional free speech construct, you were not allowed to yell “fire!” in a crowded theater — because it could cause physical harm. Now, a substantial majority of college students believe free speech means limiting speech that would cause other types of harm or cause exclusion of people or groups. The increased value of inclusion and protection from this sort of harm is intensified by common use of social media, with its reinforcement of filter-bubbles of like-minded thinkers and the ability to block anyone with whom you disagree. Anonymity, hate speech and bullying all promote the sort of thinking that values protection over exposure.

We are in the field with Gallup now and we’ll have a 2018 version of that study soon. It remains to be seen whether we have a trend or just a reflection of current events at a moment in time. If we’re right and this is a trend, it will be one of many in the ever-evolving history of the First Amendment.

The First Amendment we enjoy today, the world’s gold standard, was significantly forged in court battles over the last half-century, largely paid for by newspaper companies. Those companies either no longer exist or are financially strained, but they left a reasonably well-settled body of law affirming rights of people and press and the extension of those rights to broadcast licensees.

What is not settled are free speech rights on internet. Will courts ultimately choose freedom or restriction? Will internet speech be a right or a license from government?

As these new understandings and debates occur, we at Knight felt it important to ensure the presence of a disinterested advocate arguing for free speech. Three years ago, we partnered with Columbia University to establish the Knight First Amendment Institute at Columbia for just that purpose. With an initial endowment of $50 million, it will be an independent affiliate of the university, led by Jameel Jaffer and a team of outstanding attorneys, and guided by a board that includes Ted Olson of Gibson Dunn and Eve Burton of the Hearst Corporation, with the active support of Lee Bollinger, Columbia’s president and a leading First Amendment scholar.

The consequences are enormous and the legal questions are wide open. The Knight Institute will engage in the courts, through research, scholarship and conferences, always with a bent toward free speech and free press.

In summary, let me suggest that the ground rules going forward are both simple and hard: First, whatever we do must be in line with the First Amendment. Second, we must recognize that the problem is not monolithic, and any response must be nimble, nuanced and iterative. Third, we should remember that this challenge exists within a larger context: The massive questions regarding the decline of trust in all institutions makes this a civic emergency.

Not long ago, I spoke with a professor at MIT about the upheaval in communications technology and its consequences. I asked him where he thought we were in that revolution, on a scale of one to 10, with one being a new technology and 10 being a mature technology. He said, without hesitation, “Two, maybe three. You ain’t seen nothin’ yet.”

We are very much in the early days of the new world. After Gutenberg, society adapted to embrace his disruption and thrived as never before. I hope history repeats itself.