BAE will receive an additional $23.5 million, according to a release. That amount comes in addition to its original $47 million contract, bringing the cost of the project to over $70 million.

The company will extend the barrel of the Mk 45 to 62 calibers and strengthen the gun mounts. It will also incorporate a fully digital control system. The upgrades will allow the Mk 45s to fire more modern munitions that have 50% greater firing energy, BAE said in the release.

The work is expected to be completed by the end of 2028.

Brent Butcher, vice president of weapon systems at BAE, said the events in the Red Sea last year emphasized the need for modernized weapons systems in the Navy’s fleet. Starting in late 2023 and continuing through 2024, Iran-backed Houthi rebels launched attacks against commercial and naval ships in the Red Sea, leading to U.S. naval and air strikes against the group.

The situation “underscored the importance of firepower aboard U.S. Navy ships,” Butcher said.

“Equipped with the latest configuration of the [Mk 45] gun system, the men and women in the Navy have the capabilities to protect themselves at sea,” Butcher added. “We continue our commitment to providing the latest naval gun technology, including advanced munitions, to U.S. Sailors and their allies.”

BAE said the upgrades would come at a lower cost than introducing a new gun system altogether. The work on the Mk 45s will be completed at the company’s production facility in Louisville, Kentucky.

The Army is using its Maven Smart System to provide responders with the information needed to make quick, on-the-ground decisions, such as where to send medical supplies or how many truckloads of water to take into certain storm-ravaged areas, defense officials told reporters Monday.

Weeks after the deadly hurricane tore a path from Florida’s Gulf Coast into the Appalachian Mountains, some residents in the southeast are still sifting through the wreckage caused by floods and landslides that destroyed entire towns.

More damage is feared as Hurricane Milton bears down on Florida this week as well.

Maven is a data analysis and decision-making tool that takes in reams of data from multiple sources and uses artificial intelligence and machine learning to visualize the information.

The Pentagon originally adopted Maven to use geolocation data and satellite imagery to automatically detect potential targets on the battlefield. Its use in responding to Helene is the first instance Maven has been applied to hurricane response efforts, defense officials said.

“We can get data out of these environments that have little to no communications capabilities back into the FEMA dashboard so they understand where they need to supply things,” said one defense officials, who spoke on condition of anonymity. “I think being able to bring together this common operating picture gives us better situational awareness, helps us respond faster and facilitates getting support and supplies out quicker.”

The system eliminates the need for responders to read through spreadsheets to gather pertinent information. Instead, Maven pulls out the most important data for leaders to analyze, the official said.

The military used similar technology for disease surveillance during the Covid-19 pandemic and to track individuals during the withdrawal of U.S. forces and their allies from Afghanistan, according to officials. The operators of Maven are hoping to learn from their experience responding to Hurricane Helene, and want to hone the system for use in future natural disasters or national crises.

“That way, in an event like this, we can be part of the noble effort assisting the nation’s citizens in their most urgent time of need,” a defense official said. “We can create a platform that can be an enduring presence, ready to respond.”

Part of Maven’s job for the hurricane response is to track members of the National Guard and active-duty troops who have deployed to the areas hardest hit.

As of Monday, 7,600 troops from 18 different states had deployed to the southeast. They’re providing humanitarian relief, clearing emergency routes, assessing damaged water systems and restoring infrastructure. The Defense Department also provided hundreds of high-water vehicles and dozens of helicopters and rescue boats, Pentagon press secretary Air Force Maj. Gen. Pat Ryder said in a statement.

The U.S. Army Corps of Engineers has established 12 emergency operations centers across the southeast, including three in North Carolina alone, Ryder said.

About 7,000 federal personnel, including FEMA staff, are deployed to the area. As of Sunday, FEMA had approved $137 million in housing and other types of assistance to more than 81,500 households in North Carolina, Florida, Georgia and South Carolina. The agency also provided 15 million meals, 14 million liters of water and 157 generators.

As response efforts were underway, misinformation about FEMA’s work spread online. The high volume of false information sowed confusion among survivors and threatened response and recovery efforts, FEMA said. The agency created a webpage to try to dispel the rumors.

While the cleanup from Helene continues across the southeast, Florida was preparing Tuesday for another hurricane to make landfall this week. Hurricane Milton was expected to hit the west coast of the state on Wednesday as a Category 3, the Associated Press reported.

More than 5,000 Florida National Guard troops were mobilized Tuesday to prepare for Hurricane Milton’s arrival. Army officials moved additional personnel and equipment to Fort Moore in Georgia in anticipation of search and rescue operations.

This story was produced in partnership with Military Veterans in Journalism. Please send tips to MVJ-Tips@militarytimes.com.

An FBI investigation revealed that Teixeira, a 22-year-old who ran a server on Discord called “Thug Shaker Central,” spent much of his life online, talking primarily with other young men via message, video calls and voice chats. He chatted about guns and military gear, threatened his school, made racist and antisemitic jokes, traded conspiracy theories, discussed antigovernment sentiments, and in a bid to show off, shared some of the military’s most closely guarded secrets about the Russia-Ukraine war and the Middle East.

By the time the young airman was arrested in 2023, media scholar PS Berge had been studying Discord and its users for three years and had created an online consortium of other academic researchers who were doing the same. That an intelligence leak occurred on the site, creating a national security incident, didn’t come as a shock to her.

“My response was, ‘Of course. Of course this would happen on Discord,’” Berge said. “Because on a platform like this, you share everything with your people. Everything about your life. So, why not share national security secrets?”

Teixeira pleaded guilty in March to six counts of willful retention and transmission of national defense information. His sentencing is scheduled for September, and prosecutors are asking that he serve between 11 and 17 years in prison.

The same month Teixeira agreed to a plea deal, the FBI revealed it had investigated another service member in 2022 for leaking information on Discord.

Former Air Force Staff Sgt. Jason Gray, who served as a cyber analyst at Joint Base Elmendorf-Richardson, Alaska, admitted to running a Facebook group for followers of Boogaloo, a loosely organized, antigovernment movement that advocates for a second Civil War. Gray was disgruntled with his military career, and he discussed his dissatisfaction with the U.S. government in several Discord channels created for the Boogaloo movement, according to a 2022 FBI affidavit that was unsealed in March.

Gray, who used the account name LazyAirmen#7460, was accused of posting a classified image in a private Discord channel that he “likely obtained” from his access to National Security Agency intelligence, the affidavit states.

Investigators said the image could’ve been shared “in furtherance of the Boogaloo ideology,” but didn’t elaborate on the image’s details. It’s uncertain whether the FBI is still investigating the potential leak. But while searching Gray’s electronic devices for evidence of an intelligence breach, authorities discovered hundreds of images of child pornography. Gray is currently serving five years in federal prison on multiple child pornography charges.

Oversharing is a hallmark of Discord, an online world where members of certain channels talk all day, every day, and even fall asleep together on voice calls, said Megan Squire, a computer scientist and deputy director for data analytics at the Southern Poverty Law Center.

People who study the platform agree that it’s not inherently bad — it’s used by millions of gamers, students, teachers, professionals, hobbyists and members of the military community to communicate and socialize. However, extremists have hijacked a part of the platform to radicalize and recruit others to their causes, said Jakob Guhl, senior manager for policy and research at the Institute for Strategic Dialogue.

Following the leak of national security secrets and other high-profile, nefarious uses of the platform in recent years, researchers are grappling with what to think of the platform’s small but headline-grabbing dark side, and many disagree on whether Discord as a company is doing enough to root out bad actors.

“It’s always a bit difficult to strike the right tone between not scaring people off the platform, because the majority of users are completely fine, but also highlighting that there is an actual issue of radicalization,” Guhl said. “It’s not the biggest or most offending platform, but it definitely plays a crucial role among this network.”

‘Not inherently evil’

The National Consortium for the Study of Terrorism and Responses to Terrorism, known as START, studied decades of violent extremist attacks and found a military background to be the most commonly shared characteristic among those who committed or plotted mass casualty attacks from 1990 through 2022, more so than criminal histories or mental health problems.

Researchers from START said the study revealed why extremist groups tend to focus recruitment efforts toward people with military service records: Even a small number of them can have an outsized impact inside extremist movements.

While such recruitment occurs on Discord, Guhl, Berge and Squire agreed that the mere presence of service members and veterans on the platform isn’t a cause for concern.

“It’s a popular platform and not inherently evil,” Squire said. “I’d be much more concerned about military folks on 4chan, Telegram, places like that. Nothing good is happening on those platforms, but Discord could be useful.”

In fact, Berge said, it can be a valuable forum for marginalized people to foster a sense of community. On its “about” page, Discord describes its mission as one that helps users find a sense of belonging.

“Discord is about giving people the power to create space to find belonging in their lives,” the company’s mission statement reads. “We want to make it easier for you to talk regularly with the people you care about. We want you to build genuine relationships with your friends and communities close to home or around the world.”

That’s what the veterans group Frost Call is doing on the platform. The nonprofit encourages veterans and service members to stay connected through gaming, one of its founders told Military Times last year. As of June, it boasted 390 members.

“When we founded Frost Call, we built an organization around this idea of bringing veterans together, helping to improve camaraderie that’s missing from military service,” Marine Corps veteran Wesley Sanders said last year. “It serves an enormous mental health need, but also … an existential need for a lot of veterans.”

Moreover, when new users join Discord, extremist elements of the platform are not easily visible.

Discord is made up of millions of servers centered on various topics. Users can join up to 100 servers, and each server has numerous text, voice and video channels. When a new user creates an account and searches servers to join, the platform will suggest “its most popular, most successful, public-facing communities,” rather than any disquieting, invite-only communities, Berge said.

“If you are a standard user, and if you’re signing in to Discord for your general interests — maybe you’re looking for fellow students or fellow veterans — 90% of the time, you’re not going to accidentally stumble upon an extremist group,” she said. “They actually go through a lot of effort to make these spaces insulated, to make them difficult to find.”

When using Disboard, a third-party search platform for Discord servers, prompts such as “Nazi” or “white supremacist” won’t elicit results like they used to, Berge said. In a 2021 study, she found thousands of Discord servers that marketed themselves on Disboard as hateful and Nazi-affiliated spaces.

“You used to be able to search for those terms and find communities. It was horrifying,” Berge said. “Those servers still exist, but they’ve changed the ways they’re identified, and in some cases, we know that high-profile, toxic communities have been shut down.”

Extremists find a foothold

Founders Jason Citron and Stan Vishnevskiy created Discord in 2015 as a way to allow friends around the world to communicate while playing video games online. Its popularity exploded during the Covid-19 pandemic, when lockdowns went into effect and many people became more isolated than ever before.

Just two years after it launched, Discord gained notoriety as the platform of choice for facilitators of the 2017 “Unite the Right” rally in Charlottesville, Virginia. Organizers, including some veterans, used Discord to share propaganda and coordinate the protest, which turned deadly. James Fields was convicted of killing Heather Heyer when he drove his car into a group of counterprotesters. Fields had joined the Army in 2015 but was separated quickly because of a cited lack of motivation and failure to train.

In 2022, Discord made headlines again after a mass shooting at an Independence Day parade in Highland Park, Illinois, where seven people were killed and dozens more injured. The suspected shooter ran his own Discord server called “SS,” where he complained about “commies,” short for “communists,” according to posts archived by the nonprofit website Unicorn Riot.

That same year, an 18-year-old white gunman killed 10 Black people at a supermarket in Buffalo, New York. The gunman, Payton Gendron, spent months writing plans for the attack in a diary he kept on a private Discord server, visible only to him. About 30 minutes before the attack, Gendron sent out invitations for others to view the diary, and 15 people accessed it, according to Discord.

The platform again faced scrutiny following Teixeira’s leak of national security secrets.

“It’s periodic. Every couple of years, it seems like there’s something,” Squire said. “There are other platforms that are worse, but Discord keeps coming up over and over again.”

Research institutions such as the Institute for Strategic Dialogue found that Discord serves as a hub for socializing and community-building across far-right groups, including Catholic extremists, the white supremacist Atomwaffen Division and the antigovernment Boogaloo movement.

Extremist groups value the platform’s layers of privacy and anonymity, as well as its chat and video functions and collaborative nature, Guhl said. Berge described it as a walled garden, or an online environment where user access to content can be controlled. Servers come with the capability to assign hierarchy to different members and allow some members to access information that others can’t, the researchers said.

“In, say, a Twitter direct-messaging thread or Facebook DM, you don’t really have levels and hierarchies,” Squire said. “Discord really allows you to have more fine-grained ranking structures.”

Another reason for the prevalence of extremists on the platform stems from its roots in gaming, Guhl surmised.

Rachel Kowert, a globally recognized researcher on gaming and mental health, has spent five years researching extremism in video game communities. Though gaming itself is a powerful tool for connection and growth, extreme and hateful ideologies are now commonplace in those spaces, Kowert said.

“If you’re spending a lot of time in the social or gaming spaces where misogyny is commonplace, that can in turn start to internalize in the way you see the world and interact in it,” Kowert said.

Fighting a dark legacy

The existence of far-right groups on Discord — and the high-profile instances of extremism on the platform in the past several years — has spawned its “extremist legacy,” one from which it’s now trying hard to distance itself, said Berge.

Discord said it removed more than 2,000 far-right-affiliated servers following the “Unite the Right” rally. After the Buffalo killings, it removed Gendron’s server and worked to prevent the spread of content related to the attack, the company said. At that point, Discord agreed it “must do more to remove hate and violent extremism.”

“We created Discord to be a place for people to find belonging, and hate and violence are in direct opposition to our mission,” the company said in a statement at the time. “We take our commitment to these principles seriously and will continue to invest in and deploy resources.”

Earlier this year, the company reported that 15% of its staff works on its user safety team, which cracks down on harassment, hateful conduct, inappropriate contact, violent and abusive imagery, violent extremism, misinformation, spam, fraud, scams and other illegal behavior.

During the investigations into Teixeira and Jason Gray, Discord officials immediately cooperated with law enforcement, a company spokesperson told Military Times. And in recent months, Discord has leaned on machine-learning technology to moderate content.

“We expressly prohibit using Discord for illegal activity, which includes the unauthorized disclosure of classified documents,” the spokesperson said.

The company publishes reports each quarter showing actions taken against various accounts and servers. The latest report, published in January, says Discord disabled 6,109 accounts and removed 627 servers that espoused violent extremism during the last few months of 2023.

Squire and Guhl agreed that Discord is “pretty good” at responding to extremist content. Guhl credited the company for including extremism and hate speech in its community guidelines, as well as for deleting servers on a regular basis that breach its terms of service. Discord also created a channel where Squire could flag questionable content on the platform, and the company has been receptive to the concerns she’s raised, she said.

“I credit where credit is due, and I have to give them credit for that,” Squire said. “I think it’s taken seriously, and there are other platforms that I could not say that about.”

Extremists are ‘absolutely still there’

Berge applauded Discord for ramping up the technology behind its moderation and for introducing IP bans, which restrict a device from accessing the platform, rather than just an account. Still, she sees room for improvement.

Discord should place more emphasis on educating moderators and users about how to recognize when someone is being radicalized and pulled into an extremist space, Berge said. She also criticized the platform for disbanding a program in 2023 that included hundreds of volunteer moderators.

“It wasn’t Discord’s automated flagging systems that caught national security secrets being leaked by Jack Teixeira. It took other users and community moderators digging into it and someone finally reporting it,” Berge said. “Elevating people and giving them tools to moderate is absolutely central to protecting the platform, and that’s one area where I think they’re taking a step back.”

Berge is still researching communities on Discord, four years after she first uncovered a network of white supremacists using the platform as a recruitment ground. Despite its community guidelines and efforts to remove offending servers and accounts, Discord still serves as a meeting place for pockets of extremism.

“They’re harder to find, but they are absolutely still here. We’re still finding them,” Berge said. “It is still one of the most popular spaces for people to congregate, share and be in community with each other, for better or for worse.”

Discord remains the “platform of choice” for some hate groups, noted Squire, who described the company’s fight against extremists as playing whack-a-mole: As soon as one is removed, another pops up. A lack of institutional knowledge among far-right extremist groups is partly to blame, she said.

“Everybody’s always fresh, and they don’t have any structure for teaching one another and learning from mistakes of the past,” Squire said. “That’s convenient for us, because as we keep amassing knowledge, they make the mistake of reusing the technology that’s most convenient, rather than being strategic.”

This story was produced in partnership with Military Veterans in Journalism. Please send tips to MVJ-Tips@militarytimes.com.

Chinese actors exploited the unfolding chaos and took to social media, where they shared a conspiracy that the fire was the result of a “meteorological weapon” being tested by the U.S. Department of Defense. According to new analysis from Microsoft Corporation, a worldwide technology company, the Chinese accounts posted photos that were created with generative artificial intelligence, which uses new technology that can create images from written prompts.

The situation exemplifies two challenges that some experts are warning about ahead of the presidential election in November: The use of generative AI to create fake images and videos, and the emergence of China as an adversary that stands ready and willing to target the United States with disinformation. Academics are also voicing concerns about a proliferation of alternative news platforms, government inaction on the spread of disinformation, worsening social media moderation and increased instances of public figures inciting violence.

An environment rife for disinformation is coinciding with a year during which more than 50 countries are holding high-stakes elections. Simply put, it’s a “very precarious year for democracy,” warned Mollie Saltskog, a research fellow at The Soufan Center, a nonprofit that analyzes global security challenges. Some of the messaging meant to sow division is reaching veterans by preying on their sense of duty to the U.S., some experts warned.

“Conspiracy theories are a threat to vulnerable veterans, and they could drag your loved ones into really dark and dangerous places,” said Jacob Ware, a research fellow at the Council on Foreign Relations who published a book this year about domestic terrorism.

“The seeds are there for this kind of activity again, and we need to make the argument that protecting people from conspiracy theories is in their best interest, not just in the country’s.”

The threat of state-backed disinformation

China kept to the sidelines during the 2016 and 2020 presidential elections, watching as Russia targeted the U.S. with chaos-inducing disinformation, according to the U.S. Intelligence community. But Beijing’s disinformation capabilities have increased in recent years, and the country has proven a willingness to get involved, Saltskog said.

In a report published in April, Microsoft warned that China was preparing to sow division during the presidential campaign by using fake social media accounts to discover what issues divide Americans most.

The accounts claimed that a train derailment in Kentucky was deliberately caused by the U.S. government, and they polled Americans about issues like climate change, border policies and racial tensions. China uses the information gathered to “understand better which U.S. voter demographic supports what issue or position and which topics are the most divisive, ahead of the main phase of the U.S. presidential election,” the technology behemoth warned.

“The primary concern today, in our assessment, is Russia, China and Iran, in that order,” Saltskog said. “Specifically, when we talk about China … we’ve seen with the Hawaii wildfire example that they have AI-powered capabilities to produce disinformation campaigns targeting the U.S. It’s certainly very concerning.”

Envoys from China and the U.S. are set to meet this week to discuss the risks of artificial intelligence and how the countries could manage the technology. A senior official in President Joe Biden’s administration told reporters Friday that the agenda does not include election interference, but noted the topic might come up in discussion.

“In previous engagements, we have expressed clear concerns and warnings about any [Peoples Republic of China] activity in this space,” a senior administration official said. “If the conversation goes in a particular direction, then we will certainly continue to express those concerns and warnings about activity in that space.”

Logically, a British tech company that uses artificial intelligence to monitor disinformation around the world, has been tracking Russian-sponsored disinformation for years. Kyle Walter, the company’s head of research, believes Russia is positioned to increase its spread of falsehoods during the run-up to the November election, likely focusing on divisive issues, such as immigration and the U.S. economy.

Russia isn’t seeking to help one candidate over another, Walter added. Rather, it’s trying to sow chaos and encourage Americans to question the validity and integrity of their voting process.

Microsoft’s threat analysis center published another report at the end of April, saying that Russian influence operations to target the November election have already begun. The propaganda and disinformation campaigns are starting at a slower tempo than in 2016 and 2020, but they’re more centralized under the Russian Presidential Administration than in the past, Microsoft officials said.

So far, Russia-affiliated accounts have been focusing on undermining U.S. support for Ukraine, pushing disinformation meant to portray Ukrainian President Volodymyr Zelensky as unethical and incompetent, and arguing that any American aid to Ukraine was directly supporting a corrupt and conspiratorial regime, the report states.

State-backed disinformation campaigns like these have used images of U.S. service members and targeted troops and veterans during the previous two presidential elections.

A study from Oxford University in 2017 found Russian operatives disseminated “junk news” to veterans and service members during the 2016 election. In 2020, Vietnam Veterans of America warned that foreign adversaries were aiming disinformation at veterans and service members at a massive scale, posing a national security threat.

“It’s certainly happened historically, and it’s certainly a threat to be aware of now,” Saltskog said.

A social media ‘hellscape’

In March, the pause in public sightings of Kate Middleton, along with the lack of updates regarding her health following abdominal surgery, created a breeding ground for conspiracy theories.

Scores of memes and rumors spread in an online fervor until Middleton posted a video March 20, in which the Princess of Wales shared the news that she had been diagnosed with an undisclosed type of cancer. Some who shared conspiracies responded with regret.

The situation offered a public example of how conspiracy thinking snowballs on social media platforms, as well as the real harm it can cause, said AJ Bauer, an assistant professor at the University of Alabama, where he studies partisan media and political communications.

“It does reinforce the fact that social media gives us an opportunity to crowdsource and spin conspiracy theories and conspiracy thinking,” Bauer said. “It can be done for a kind of whimsy, or it can be done for harm, and the line between whimsy and harm is a fine one. It can tip from one to the other pretty quickly.”

Social media platforms were blamed during the presidential election seasons in 2016 and 2020 for the influential campaigns that gained traction on their sites. Since then, the situation has only worsened, several experts argued.

Bauer specifically blamed recent changes at X, formerly Twitter, which was taken over in 2022 by billionaire Elon Musk.

Musk has used the platform to endorse at least one antisemitic conspiracy theory, and several large corporations withdrew from the platform after their ads were displayed alongside pro-Nazi content. Bauer described the site as a “hellscape” that is “objectively worse than it was in 2020 and 2016.”

“One really stark difference is that in 2020, you had a lot of big social media platforms like Twitter and Meta trying at least to mitigate disinformation and extremist views because they got a lot of blame for the chaos around the 2016 election,” Bauer said. “Part of what you see going into this election is that those guardrails are down, and we’re going to experience a free-for-all.”

In addition to the changes at X, layoffs last year struck content-moderation teams at Meta, Amazon, and Alphabet (the owner of YouTube), and have led to fears that the platforms would not be able to curb online abuse or remove deceptive disinformation.

“That timing could not be worse,” Saltskog said. “We’re going to have a much thinner bench when it comes to the people inside social media platforms who do this.”

Kurt Braddock, an assistant professor at American University who studies extremist propaganda, argued that social media platforms don’t have financial incentive to moderate divisive or misleading content.

“Their profits are driven by engagement, and engagement is often driven by outrage,” Braddock said.

Because of this, Braddock believes there should be more focus placed on teaching people — especially younger people — how to spot disinformation online.

A study last year by the Center for Countering Digital Hate found that teenagers were more likely than U.S. adults to believe conspiracy theories. Another study by the University of Cambridge found that people ages 18 to 29 were worse than older adults at identifying false headlines, and the more time people spent online recreationally, the less likely they were able to spot misinformation.

The average age of the active-duty military is 28.5, according to a Defense Department demographic profile published in 2023, and new recruits are typically in their early 20s. Because of the young age of the force, Braddock thinks the Defense Department should be involved in teaching news literacy.

The Pentagon and Department of Veterans Affairs did not respond to a request for comment about any efforts to help service members and veterans distinguish accurate information online.

“They’ve grown up in the digital age, and for some it’s been impossible to differentiate what’s real and what’s not real,” Braddock said. “They’ve essentially been thrown to the wolves and don’t have the education to be able to distinguish the two. I think there needs to be a larger effort toward widespread media literacy for young people, especially in populations like the military.”

AI speeds disinformation

Overall, people are getting better at spotting disinformation because of awareness efforts over the past several years, Braddock believes.

However, just as more people were becoming accustomed to identifying false information, the landscape changed, he said. Generative AI gained traction last year, prompting the launch of tools that can create new images and videos, called deepfakes, from written descriptions.

Since Hamas launched a surprise attack on Israel on October 7, these digital tools have been used to create propaganda about the ensuing conflict. Some AI images portray injured or frightened Palestinian children running from air strikes in Gaza, while others depict crowds of people waving Israeli flags and cheering for the Israeli Defense Forces.

“This is a new thing, and most people aren’t prepared to differentiate between what’s real and what’s not,” Braddock said. “We need to stay on top of these different technologies which can be used to reach a large amount of people with propaganda. It only takes one person to do a lot of damage.”

The Soufan Center and Council on Foreign Relations consider AI to be the top concern heading into November. The technology is developing faster than Congress can work to regulate it, and social media companies are likely to struggle to moderate AI-generated content leading up to the election, Ware said.

As Congress grapples with reining in AI, some states are taking the lead and scrambling to pass legislation of their own. New Mexico is the latest to pass a law that will require political campaigns to provide a clear disclaimer when they use AI in their advertisements. California, Texas, Washington, Minnesota and Michigan have passed similar laws.

Ware said that among those working to counter domestic terrorism, AI has been treated like “a can getting kicked down the road” — a known problem that’s been put off repeatedly.

“We knew it was a specter that was out there that would one day really affect this space,” Ware said. “It’s arrived, and we’ve all been caught unprepared.”

The technology can accelerate the speed and scale at which people can spread conspiracy theories that radicalize others into extremist beliefs, Ware added. At The Soufan Center, researchers have found that generative AI can produce high quantities of false content, which the center warned could have “significant offline implications” if it’s used to call for violent action.

“You can create more false information at scale. You can just ask an AI-powered language model, ‘Hey, create X number of false narratives about this topic,’ and it will generate it for you,” Saltskog said. “Just the speed and scale and efficacy of it is very, deeply concerning to us experts working in this field.”

The creation of AI images and videos is even more concerning because people tend to believe what they can see, Saltskog said. She suggested people look at images and videos carefully for telltale signs of digital deception.

The technology is still developing, and it’s not perfect, she said. Some signs of deepfakes could be hands that have too many or too few fingers, blurry spots, the foreground melding into the background and speech not aligning with how the subject’s mouth is moving.

“These are things the human brain catches onto. You’re aware of it and attune to it,” Saltskog said. “Your brain will say, ‘Something is off with this video.’”

As Congress and social media platforms lag to regulate AI and moderate disinformation, Americans have been left to figure it out for themselves, Ware argued. Bauer suggested people do their homework about their source of news, which includes determining who published it, when it was published and what agenda the publisher might have. Saltskog advised people to be wary of anything that elicits a strong emotion because that’s the goal of those pushing propaganda.

Similarly, Ware recommended that if social media users see something that seems unbelievable, it likely is. He suggested they look for other sources providing that same information to help determine if it’s true.

“People are going to take it upon themselves to figure this out, and it’s going to be through digital literacy and having faith in your fellow Americans,” Ware said. “The stories that are trying to anger you or divide you are probably doing so with an angle, as opposed to a pursuit of the truth.”

This story was produced in partnership with Military Veterans in Journalism. Please send tips to MVJ-Tips@militarytimes.com.