Kevin Cohen, the CEO and co-founder of Tel Aviv-based startup RealEye, opens his computer, ready to introduce the tools that he believes will keep America’s borders more secure.
The screen fills with photos of people young and old, from all ethnicities, genders, and backgrounds. “We’re basically aggregating information from hundreds of thousands of sources online, not only to classify someone as a potential danger or threat, but also to indicate the probability of them meshing into various nefarious networks,” says Cohen. “We can take 50,000 individuals and within literally seconds, I can give you the insights and the indexing of their behavior.”
RealEye has developed two AI-driven platforms, Masad and Fortress, that provide real-time vetting and “semi-active” monitoring of immigrants entering a country either legally or otherwise, with a history of illegal or suspicious activities. These tools draw not just on criminal records (which could be as innocuous as traffic violations), but also digital footprints left on social media and the dark web.
Cohen pulls up the profile of a man named Yosef, and instantly the man’s history appears like bullet points on the screen. “We can see that this specific individual funnels money to Hamas and other operatives,” says Cohen. “We know this guy is no good.”
In just a few days, Donald Trump will be inaugurated for his second term as US president. He’s vowed that on Day One, his priority will be “the largest deportation operation in American history,” and there is “no price tag.” Trump’s plans to secure our Southern border include everything from a proposed hotline center for citizens to call in tips on undocumented migrants to building a massive immigration detention center on a 1,402-acre plot on the Rio Grande. Trump has also indicated his intention to continue construction of a 30-foot-high fence across the US-Mexico border. In fact, he’s considering declaring illegal immigration a national emergency to unlock funds for border wall construction.
In December, Trump even filed an amicus brief in support of a legal effort to stop the Biden administration from selling border wall materials. A court order put a stop to Biden’s efforts. Further emboldened, Trump will likely also revisit plans he’d considered in his first term, like floating barriers, which were briefly tested in the final year of his presidency. In 2023, Texas Gov. Greg Abbott picked up on the idea, constructing 1,000 feet of sphere buoys in a section of the Rio Grande. Despite attempts by the Biden administration to stop it, claiming the floating barriers violated the federal Rivers and Harbor Act, the Fifth Circuit Court of Appeals ruled last summer they could remain.
Along with these physical barrier strategies, Artificial Intelligence-based strategies such as those found with RealEye will likely form a large part of both securing our Southern Border and Trump’s deportation strategy. The Department of Homeland Security has already been allocated $5 million in its 2025 budget to open an AI office, and DHS Secretary Alejandro Mayorkas has called AI a “transformative technology.” Sam Altman, the OpenAI CEO who donated $1 million to Trump’s inauguration fund, predicted that President Trump “will lead our country into the age of AI.”
Another AI company preparing for a busy year is Bavovna AI, a Ukrainian startup who’ve developed autonomous, AI-powered drones. Although drones have been used on the US border since 2005, they’ve been increasingly sabotaged by Mexican cartels with GPS-jammers. But Bavovna’s drones — already battle-tested in Ukraine’s war with Russia — can traverse complex landscapes without GPS or any visual odometry, thus evading detection.
Like RealEye’s Cohen, Bavovna’s co-founder and director Pramax Prasolov, is tight-lipped about their relationship with the upcoming administration, other than saying they’re “exploring opportunities to support border security operations.” But on Christmas Eve, Prasolov hosted a special demo flight test in Tampa, Florida, whose guests included high-ranking officials from the Department of Defense and the US.Air Force.
“These UAVs are not just tools for surveillance but solutions for maintaining operational superiority in contested environments,” Prasolov says. It’s only a matter of time before “electronic warfare” becomes “essential to secure U.S. borders,” he says.
AI is nothing new to the border; during Biden’s presidency, Democrats pushed for “smart border” technologies like virtual towers, which promised a more “humane” alternative to Trump’s border wall. And last fall, the bipartisan Emerging Innovative Border Technologies Act was passed by the House compelling the DHS to explore how to best utilize AI, machine-learning and nanotechnologies in border security programs.
The Trump administration clearly appears on board. They “plan to increase the use of AI surveillance systems along the border,” says N. And Trump is taking off the safety bumpers. Biden signed an Executive Order in 2023, vowing a “safe, secure, and trustworthy” development of AI; Trump has promised to nullify it when he takes office.
There are plans already in place to improve on the existing AI, like a proposed $101 million to upgrade and maintain surveillance towers. “These towers . . . detect and identify unauthorized crossings, and distinguish between humans, animals, and vehicles,” says Sahota. “Under a Trump administration, this initiative is expected to expand radically, turning his promise of a physical wall into a digital one.” In fact, Customs and Border Protection plans to complete a network of more than 1,000 surveillance towers by 2034.
There will also be a full introduction of “robodogs,” first unveiled by the DHS in 2022. These four-legged military-grade machines, developed by Philadelphia-based Ghost Robotics, “are designed to navigate challenging terrains, including sand, rocks, and human-made environments like stairs, making them well-suited for the diverse landscapes encountered along the border,” says Sahota. Although they’re currently still being tested by the DHS, Trump is enthusiastic enough about the technology to have at least one robodog on security patrol at Mar-a-Lago.
There have been concerns that the dogs could be weaponized, especially after some models were introduced with sniper rifles. The last company that attempted something similar, tech startup Brinc, faced public backlash in 2021 over drones they manufactured equipped with tasers. But Ghost Robotics CEO Gavin Kenneally assured a House Oversight Committee in 2023 that their technology would only be used for patrolling, not violence. (Ghost Robotics did not respond to requests for comment.)
Some of the other AI being introduced isn’t quite so dystopian, but it’s no less complex. RealEye’s vetting system, for instance, goes beyond existing DHS strategies, which focus mostly on biometric indexing. This is insufficient, says Cohen, to intercept bad actors — such as terrorists or cartel members — who are exploiting the migration system to enter the US illegally. “You have somebody’s fingerprint and that’s it,” explains Cohen, who says his firm has partnered with intelligence agencies in Israel, Taiwan, the Philippines and Japan. “There’s no tracking of their digital presence…[to gauge] who these individuals are before they came into the country.”
Machine-learning technologies are also far better suited to identify patterns and anomalies that might be overlooked by border patrol agents — overwhelmed by the surge in Biden-era migrants and whose current staffing crunch has been called “unsustainable” by a DHS report in 2023. “Some AI systems…can analyze communication intercepts in multiple languages to identify potential security threats,” says Sahota. “AI can also…identify…unusual movement patterns near the border or irregular travel itineraries can be flagged for further investigation.”
Although many of these tools are still being field-tested, there have already been success stories. In 2023, an AI model identified the unusual travel patterns of a truck making regular crossings at the U.S./Mexico border, and when police investigated, they discovered 75 kilograms of narcotics hidden in the vehicle.
But the rise of AI in border security isn’t giving everyone comfort. Sahota calls it a double-edged sword. “While AI offers vast potential to streamline operations and secure borders, its misuse—intentional or not—could lead to dangerous outcomes,” he says. “There’s a slippery slope where AI can be weaponized, eroding civil liberties, violating privacy, and fostering systemic bias. Under the Trump administration, the aggressive expansion of AI in border security could turn these tools into mechanisms for mass surveillance and control.”
Marina Shepelsky, the CEO, co-founder, and immigration attorney at New York-based Shepelsky Law Group, worries that AI vetting could have built-in biases. “They can vet and deny people who have criticized Trump,” she says. “Or anyone who has posted some religious Muslim thoughts, or anyone who has posted ‘communist’ thoughts on social media.”
Cohen, however, thinks concerns about AI being abused are misdirected. “The ‘special interest aliens’ we’re going after are technically felons,” he says. “We’re not targeting individuals that are just trying to become breadwinners for their family or make a better life.” For Cohen, making life difficult for immigrants with criminal intentions is very much personal. As a teenager, his family briefly moved from Israel to Florida, and they lived next door to Mohammed Atta, who would become a ringleader of the September 11 attacks.
“I’ve never stopped thinking about it,” he says. “That’s why we’re doing this, to keep guys like him . . . from making it into the country. We have to be smarter than them, and find more creative ways to make sure everybody crossing the border is who they say they are.”
This post was originally published on here
