SAN DIEGO—Verance, a global watermarking company that powers AI provenance and broadband features for broadcast television, is launching a new solution based on open watermarking standards that will help companies combat misinformation and and AI-generated disinformation.

The solution enables social media platforms to authenticate broadcast content in user uploads, enhancing public trust in media and reducing the societal risks of disinformation. It uses Verance Aspect to enable broadcast content to be reliably authenticated by social media platforms, devices and disinformation analysts.

"These tools we are providing to enable reliable authentication of broadcast news content through open standards like ATSC watermarking and C2PA provenance are essential,” explained Nil Shah, CEO of Verance. “They represent a significant step toward maintaining public trust and ensuring the integrity of the information that shapes our world in this contentious election year and beyond."

The technology stems from fully open watermarking standards developed and published by the U.S.‑based broadcast standards organization ATSC and the provenance authentication technology standards consortium C2PA. Each standards organization has broad participation across the global media, technology and consumer electronics sectors.

Using these industry standards, broadcast adopters of Verance Aspect can enable recipients of their content to determine whether it has been modified, as well as to locate an authentic version of the published asset. This validation step, which can be fully automated, allows the recipient to determine whether the media seen misrepresents what was originally broadcast, Verance reported.

Major broadcasters across the U.S., including Graham Media Group, Gray Television, Sinclair Broadcast Group and Capitol Broadcasting Company have already deployed ATSC watermarking using Verance Aspect. ATSC watermarking has also been incorporated into the DVB and HbbTV interactive television specifications, which are used in dozens of countries across Europe, Asia, Africa and the Pacific Islands.

Additionally, C2PA is beginning to be adopted as a means of authenticating content on social media platforms, with LinkedIn and TikTok having recently announced plans to use it to authenticate media uploaded by users.

Verance Aspect employs ATSC watermarking to access C2PA manifest services for broadcast content. C2PA authentication requires signed metadata that cannot be transmitted through existing content distribution paths. ATSC watermarking enables recipients of broadcast content to locate the absent C2PA metadata by embedding imperceptible URLs throughout the broadcast audio and video that survives any distribution path.

The ATSC watermark standards have supported metadata-driven solutions like C2PA since their inception in 2017. The recent release of C2PA specifications (v2.1) introduces support for watermark-based recovery of C2PA metadata and identifies ATSC watermarking as a compatible technology.

This standards-based approach enables broadcasters and platforms to create interoperable implementations of the underlying technologies, encouraging global cooperation on addressing disinformation challenges accelerated by rapid deployment of powerful generative AI systems.

A number of broadcasters applauded the initiative. "Concern about deep fakes in the media is a growing concern for the public and political leaders,” said Anne Schelle, managing director of Pearl TV. “Good news: the broadcast industry proactively addressed this issue years ago and it’s being implemented now. The NextGen TV ecosystem provides consumers with security assurance for over-the-air television broadcasts. Extending that assurance to social media and web browsing using ATSC watermarking and C2PA provenance is an essential next step to maintain consumer trust in broadcast news."

"At Gray, we believe in the power of trustworthy journalism,” David Burke, senior vice president and CTO of Gray Television, added. “By integrating Verance Aspect watermarking with C2PA standards, we're not just protecting our content; we're reinforcing the credibility of broadcast news in an age where AI-driven misinformation is a significant threat. This technology is an important part of ensuring the public receives verified, unaltered information."

"As AI continues to evolve, so too must our strategies for safeguarding the truth. This collaboration is significant as the industry battles against disinformation," said Christina Hartman, vice president and head of Scripps News for the E.W. Scripps Company. Scripps works with Verance through its relationship with Pearl.

Mike Kralec, senior vice president and CTO for Sinclair, said, "Ensuring our viewers have the highest confidence in our programming is critical to our business. We believe C2PA, through its ability to authenticate the name of content creators or distribution services, can play a central role in future end-to-end technical architectures and workflows. We applaud C2PA and Verance for recognizing that authentication of creators and distributors is absolutely essential to the advanced workflows our industry will rely on in the future."

"We are pleased to have the opportunity to cooperate with C2PA and our membership on extending the use of ATSC standards to ensure that broadcast content can be presented to viewers with an assurance of authenticity," added ATSC president Madeleine Noland.

WASHINGTON, D.C.—The Federal Communications Commission has announced that it will move forward on a previously announced proposal to implement new requirements to disclose AI-created content in TV and radio ads.   

The plan drew immediate opposition from both Republican-appointed commissioners, commissioner Nathan Simington and commissioner Brendan Carr. 

The FCC said the proposal aims to increase transparency by having those who already have legal duties to file information about their TV and radio advertisements with the FCC to indicate if AI is being used and make on-air disclosure of AI use. It does not propose the prohibition of such content, only disclosing the use of AI within political ads, the FCC stressed.  

“Today the FCC takes a major step to guard against AI being used by bad actors to spread chaos and confusion in our elections.  We propose that political advertisements that run on television and radio should disclose whether AI is being used,” said chairwoman Jessica Rosenworcel.  “There’s too much potential for AI to manipulate voices and images in political advertising to do nothing. If a candidate or issue campaign used AI to create an ad, the public has a right to know.”

In announcing the release of a Notice of Proposed Rulemaking that must still be adopted by the FCC before disclosure requirements could be implemented, the FCC stressed that artificial Intelligence has become powerful enough to mimic human voices and create life-like images. This year in the primary election in New Hampshire, thousands of voters got an AI-generated robocall impersonating President Biden that told them not to vote. This past summer, the campaign of Governor DeSantis was flagged for circulating fake AI-altered images of former President Trump.  Facing a rising tide of disinformation, roughly three-quarters of Americans say they are concerned about misleading AI-generated content, the FCC reported. 

The FCC also argued that Congress has granted the FCC authority regarding the political messages people see on television, hear over the radio, or receive over the phone.  Since the 1930s, the FCC has used this authority to require broadcasters to maintain a publicly available file for political ads.  This file has information about who bought a campaign ad, how much they paid for it, and when it ran. Over time Congress expanded these requirements to include ads that run on cable and satellite. These are also the policies that led to what are now familiar on-air disclosures so that every viewer and listener knows who is responsible for every ad. 

The FCC also stressed that the Federal Elections Commission is considering a separate rulemaking on AI, announcing this year that they expect to act in early summer.  In a recent letter to the FCC, the FEC’s Vice Chair wrote, “No one agency currently has the jurisdiction or the capacity to address every aspect of this large and complicated issue.” While the FEC can regulate AI use in online advertisements for federal candidates, the FCC can focus on the areas where the FEC is not able to act.  The FEC does not oversee television and radio stations.  Under the law, FEC authority over campaigns is limited to federal political candidates and does not extend to independent issue campaigns or state and local elections. 

Nearly half of states across the country have enacted laws to regulate the use of AI and deepfake technology in elections and most of these laws are bipartisan, the FCC also said. This FCC proposal would seek to bring uniformity and stability to this patchwork of state laws, seeking to bring greater transparency in our elections.

When Rosenworcel first announced that the FCC should consider proposals to disclose AI content in political ads, the idea quickly attacked by FCC Commissioner Carr as unworkable, beyond the agency’s authority and counter productive. 

Carr, who wrote a chapter in Project 2025 outlining plans for the FCC if Donald Trump is elected president, continued his opposition to the idea in a July 25 statement, saying it was part of a larger plan by the Democratic Party to regulate and limit political speech: “The Democratic National Committee (DNC) is now working to change the rules of the road in the run-up to the 2024 election. It has done so by calling on the administrative state to impose new controls on political speech before voters hit the ballot boxes this fall. The FCC proposal adopted today echoes that DNC-backed initiative and would impose new regulations on the use of AI-generated political speech at the eleventh hour. This push for new regulations comes on the heels of press reports that the DNC and Democrat candidates are `nervous about not keeping up with the GOP in embracing” artificial intelligence technologies in this election cycle.’”

Carr also argued that “Congress has not given the FCC the type of freewheeling authority over these issues that would be necessary to turn this plan into law.” 

He called the proposal “a recipe for chaos” because it would only apply to TV and regulatory ads because the FCC lacks the authority to regulate ads on the internet, social media and other digital platforms.”

The proposed rules, he added, are ill advised in the runup to a major election because it would mean that “For starters, the FCC is wading into an area rife with politicization. It is not difficult to see how partisan interests might weaponize the FCC’s rules during an election season.”

Nor, would it do much to solve the problem of political ads that confuse voters, he contended. 

“The FCC’s involvement can only amplify that confusion,” Carr wrote. “The FCC is legally powerless to adopt uniform rules in a technologically neutral fashion. Its legal authority claimed here extends only to legacy media—broadcast, cable, and satellite television—but not over-the-top content, like online streaming video and social media, where millions of Americans see political ads.  As a result, AI-generated political ads that run on traditional TV and radio will come with a government-mandated disclaimer but the exact same ad that runs on a streaming service or social media site would not.”

“I don’t see how this type of conflicting patchwork could end well,” Carr concluded. “Consumers don’t think about the content they consume through the lens of regulatory silos. They just view content on screens. Will they conclude that the absence of a government warning on an online ad means that the content must be real? And applying new regulations on the broadcasters the FCC regulates but not on their largely unregulated online competitors only exacerbates regulatory asymmetries.”  

MIAMI—TelevisaUnivision has announced that Univision Noticias’ fact-checking program, elDetector, will partner with Meta on fact-checking by joining Meta's Third-Party Fact-Checking Program.

Univision Noticias launched elDetector in 2016 as the first fact-checking platform in Spanish in the United States. It was later relaunched, during the 2020 presidential campaign, with support from WhatsApp and coordination by the International Fact-Checking Network from the Poynter Institute, as part of the FactChat project, which brought together 12 fact-checking platforms in the United States.

“With elDetector being the first-fact checking platform available in Spanish for the United States, our partnership with Meta will expand our reach in ensuring that Hispanic audiences receive accurate and reliable information,” said Carlos Chirinos, director of elDetector.

Users can access elDetector’s services via the Univision Noticias website, to find checked information about politics, health, and general affairs, from Covid-19 and vaccines to U.S. elections and the war in Ukraine. It also clarifies fake news and conspiracy theories, which will help users clear through the clutter of disinformation and misinformation. 

Spanish speakers can also ask questions directly to the news-checking team by sending text messages via WhatsApp to the number +1(786)685-8284. 

“This partnership emphasizes Univision Noticias’ core value of ensuring that Hispanic audiences are accurately informed and can rely on a news division to enforce their equitable access to correct and trustworthy information,” added José Gonzalo, senior director and editor in chief of Univision Noticias Digital.

“We’re thrilled to welcome TelevisaUnivision into our global fact-checking program,” said Kaitlin McCulley, news integrity program manager at Meta. “We partner with more than 90 fact-checking organizations around the world who review and rate viral misinformation on our platform, and we’re excited to build on our previous partnerships with TelevisaUnivision to help connect Spanish-speaking audiences in the U.S. with reliable information.”

In addition to this partnership, elDetector has sponsorships in coordination with the International Fact-Checking Network to do climate change checks and from WhatsApp to operate the chatbot.

We learned long ago that technology is a two-way street. History has shown that the greatest technological innovation of each era always has a dark side. Gamechanging inventions such as the automobile, personal computer, smartphone and the internet proved this to be true.

Today, our networked culture faces a serious new challenge—fake news or disinformation. Yes, the era of Walter Cronkite is definitely over. Today, you cannot believe what you see or read. 

We are now living in the “Wild West” of disinformation, where anyone can say anything, no matter how outrageous or untrue. The saddest part is too many people believe this stuff. That was proven at the U.S. Capitol on Jan. 6, 2021.

When Tim Berners-Lee invented the World Wide Web in 1989, his invention was heralded as an opportunity for anyone on earth to have a global library at their fingertips. It did that, but—as with most technology—it had another side. In addition to changing our culture, the World Wide Web became a vast wasteland of ads and a conduit to disinformation representing the worst of human nature.

Original or Fake?
Many think it is too late for the internet. The cat is out of the bag, so to speak, and we’ll never go back to a more united, cohesive society. However, some scientists are now trying to fix what the internet wrought. It won’t be easy now that anyone can acquire cheap tools to build deep fake video and photos.

One international project trying to fix the fakery problem—called "Dissimilar"—is at the Open University of Catalonia (UOC) in Barcelona, Spain. It also includes academics from the Warsaw University of Technology (Poland) and Okayama University (Japan). The researchers are working on ways to differentiate original and fake multimedia content by combining techniques from digital forensics analysis, watermarking and artificial intelligence.

“The project has two objectives: first, to provide content creators with tools to watermark their creations, thus making any modification easily detectable; and second, to offer social media users tools based on latest-generation signal processing and machine-learning methods to detect fake digital content,” said David Megías, a lead researcher.

It’s a great irony that the same artificial intelligence that brought deep fakes to the world is now being used in an attempt to eradicate them. 

We know that the rise of social media goes hand-in-hand with the increase in disinformation. Artificial intelligence made it easier to tamper with videos and photos, turning them into fakes so good that most people cannot tell the difference from the real thing.

When video and still images are combined and superimposed using artificial intelligence with advanced voice technology, the montages can look and sound like the genuine person or thing. Using this technology, known public figures can be made to say outrageous things.

Public Trust
People of different cultures perceive information in unique ways. These perceptions are being gauged by the researchers in a range of places and cultural contexts to incorporate individual idiosyncrasies when designing the solutions. 

“This is important because, for example, each country has governments and/or public authorities with greater or lesser degrees of credibility,” said Andrea Rosales, a UOC researcher. “This has an impact on how news is followed and support for fake news. If I don’t believe in the word of the authorities, why should I pay any attention to the news coming from these sources? 

“This could be seen during the Covid-19 crisis: in countries in which there was less trust in the public authorities, there was less respect for suggestions and rules on the handling of the pandemic and vaccination,” Rosales added.

A problem is people can watermark and protect disinformation as well as truthful information. I can easily see these disinformation-prevention tools being used to protect fake news from fraudulent players.

Sadly, we live in a society where huge numbers of people have developed into “disinformation cults” where facts don’t matter. They choose to block out information they don’t believe. This is a result of the internet creating niches of personal interest. What was originally conceived as a widening of available accurate information has resulted in distortions of the facts. 

Cheap artificial intelligence tools give huge power to those who would abuse them. This abuse now happens routinely, and it will be hard to stop. Prevention tools like those being developed at UOC can certainly help legitimate content providers, but they won’t stop disinformation if the masses believe it.

Dan Gillmore, who teaches at the Arizona State University Walter Cronkite School of Journalism and Mass Communication, advises readers to “create an internal speedbump.”

“Say to yourself… ‘Just wait a minute,’” before believing anything on the internet, said Gillmore. “Skepticism, especially with highly sensational titles, is key.” He advises people to corroborate stories before believing them as fact.

Developing these critical evaluation skills, however, will have to come from good media education, which is very rare in today’s schools. Only when users of the internet are taught about the realities of disinformation and how to detect it, can we begin to tackle the problem.

“Information is only as reliable as the people who are receiving it,” said Julia Koller, a learning developer. “If readers do not change or improve their ability to seek out and identify reliable information sources, the information environment will not improve.”  

WASHINGTON—NAB, E.W. Scripps, IT consultancy Fincons Group and the News Literacy Project will demonstrate how local TV newsrooms can use NextGen TV’s dynamic, interactive capability to combat fake news.

“The important role broadcasters play in educating and engaging American viewers is more vital than ever as we look at the landscape of misinformation that has permeated our culture,” said NAB CTO Sam Matheny.

The demonstration, part of the NAB Show Premiere session on April 21 at 12:30 p.m. EDT, will examine how these capabilities can be used to improve the news literacy of the public and combat misinformation, which NAB calls “one of the biggest problems facing our democracy—and the media industry—today.”

During the eight-minute demo, real-life examples of how consumers can research credible news sources and test their own news literacy skills by engaging with interactive content via a NextGen TV-compliant application that offers quizzes, polls, tips, a VOD library and other resources that can be used during live programming, NAB said. 

“NextGen TV is a terrific opportunity for companies like Scripps to further our mission to inform, engage and entertain our consumers,” said Scripps President and CEO Adam Symson. 

An interactive Q&A and networking session with tech, journalism and news literacy experts from the companies participating in the demo will take place at 1 p.m. 

“We’re excited that news literacy is the focus of this demonstration that shows how new technology can be used in innovative ways to educate people to recognize fact from fiction,” said Darragh Worland, vice president of creative services for the News Literacy Project.

NEW YORK—In this heightened age of “fake news,” Univision has announced that it is partnering with Luminate and the International Fact-Checking Network to host a workshop during the seventh edition of Global Fact.

The workshop is titled “¿Qué chequear, cómo chequear y con qué herramientas lo hago?” (“What to check, how to check and what tools do I use to do so?”), and it is meant to train Hispanic journalists in the U.S. tricks for information verification in Spanish so as to not spread news that is false or manipulated.

Cristina Tardáguila, IFCN associate director, will teach the two-part workshop. Tamoa Calzadilla from Univision will serve as the moderator.

The workshop will take place on June 29 and June 30. Find out more and register online.

Localism is a key factor in American's trust of the news media, and local TV leads the way.

That is according to 2018 Poynter Media Trust Survey and comes as the President launches almost daily attacks on the press as enemies of the people and conspirators with his opponents, and as the threat of "real" fake news from foreign actors has Washington forces on edge and critically eyeing edge providers.

A large majority (76%) of Americans have either a "great deal" or "fair amount" of trust in local TV news, followed closely by local newspapers, the study found. But only 59% say that about national newspapers, 55% for network news and 47% for online news.

The study found that, generally, only 12% trust the press a great deal while 44% say they have a fair amount of trust.

Only a third of the respondents (32%) said the media deal fairly with both sides.

That's the bad news. But there was good news as well.

[Read: Survey: Local TV News Is Booming]

The good news is that those levels of trust and confidence are actually higher than a 2017 survey. In 2016, for example, only 32% said they had either a great deal of fair amount of trust in the media. For the current survey, that total is 54%.

In addition, a strong majority (66%) says news organizations “keep political leaders from doing things that shouldn’t be done."

The survey found that more than 25% of Americans say the government should have the right to revoke the broadcast licenses of news outlets that fabricate stories. The FCC already has that power over deception from broadcast licensees who are licensed to operate in the public interest. But the issue became a political flashpoint with President Donald Trump suggested licenses should be revoked over stories he believed were made up, a category that appears to also include stories critical of him and/or using unnamed Administration sources.

FCC chair Ajit Pai has pledged to Congress there would be no politically-driven license revocations.

According to Poynter, the study was conducted "among a representative sample of the U.S. population by the survey company YouGov, which recruits a large panel of opt-in respondents and then uses a weighting and matching algorithm to construct a final sample that mirrors the demographic composition of the U.S. population."

News in general, but television news in particular, has been on the giving and receiving end of “fake news” accusations at least since the 2016 campaign season.

Even if the media or politicians were partaking in this nefarious manipulation of public opinion, those who subscribe to the notion that the truth ultimately wins out in a free market of ideas could take some solace in the belief that reality is eventually revealed through seemingly endless rounds of reporting, analysis and debate.

That’s all well and good in a Fake News 1.0 world where the battle of ideas isn’t skewed by undetectable gaming of the process.

However, I fear that A.I. technology is about to usher in Fake News 2.0, an era in which it will be impossible to believe one’s own ears and eyes. Look no further than LyreBird, an A.I. app that in the words of the company “allows you to create a digital voice that sounds like you [or anyone else, for that matter] with only one minute of audio.” If you are unfamiliar with LyreBird, you might want to check out https://youtu.be/YfU_sWHT8mo.

Now sprinkle in a couple of other A.I. developments. The first is a paper presented at SIGGRAPH 2017 entitled “Audio-Driven Facial Animation by Joint End-to-End Learning of Pose and Emotion,” which describes the use of machine learning to drive a real-time 3D facial animation from an audio input with low latency. 

The second is a paper by Berkeley A.I. Research (BAIR) researchers entitled “Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks.” The paper describes “an approach for learning to translate an image from a source domain X to a target domain Y in the absence of paired examples.”

Jun-Yan Zhu, the lead author of the paper, has posted a short YouTube video of this approach being used to transform a horse into a zebra.

The obvious danger is that some ambitious reporter one day will use perfected versions of these technologies to create the A.I.-based equivalent of “Jimmy’s World.”

However, it seems to me that the less obvious, but even more pernicious effect will be the use of these technologies to mislead reporters and editors by government entities, industry and others with an ax to grind or reason to deceive.

As one who has spent years reporting on emerging technologies that TV broadcasters could use to better inform the public –Ku-band SNG, IP newsgathering, dual polarization weather radar and mobile tech to name a few– I am truly saddened that the next wave in technology will make it extremely easy to lie to the public.

How long will television news endure when viewers no longer can believe their own eyes and ears?

SEATTLE—Researchers at the University of Washington have created video of President Obama saying something he did not say. Using the former president as a subject, Supasorn Suwajanakorn, Steven M. Seitz and Ira Kemelmacher-Shlizerman synthesized a “high-quality video of him speaking with accurate lip sync, composited into a target video clip.”

They described the process as follows in their article, “Synthesizing Obama: Learning Lip Sync from Audio,” published in this month in ACM Transactions on Graphics:

“Trained on many hours of his weekly address footage, a recurrent neural network learns the mapping from raw audio features to mouth shapes. Given the mouth shape at each time instant, we synthesize high-quality mouth texture, and composite it with proper 3D pose matching to change what he appears to be saying in a target video to match the input audio track. Our approach produces photorealistic results.”

The team said they selected Obama for several reasons, including an “abundance” of publicly available high-definition video footage from his weekly addresses comprising 17 hours and 2 million frames over eight years, plus the consistency of lighting and other production variables as well as the former president’ typically “serious and direct tone.”

Creating an otherwise realistic lip-sync “due in part to the technical challenge of mapping from a one-dimensional signal to a 3D time-varying image, but also due to the fact that humans are extremely attuned to subtle details in the mouth region; many previous attempts at simulating talking heads have produced results that look uncanny.”

Implications for actual fake news notwithstanding, the researchers said the technology could have beneficial applications in the way of reducing transmission bandwidth, creating video synthesis that enables lip-reading from audio, and affecting avatars in special effects and games.

The researchers also prepared a presentation for SIGGRAPH 2017.