Whether captured in declassified military footage or in smartphone videos uploaded to social media, UFOs are swarming Earth’s skies and demonstrating capabilities so astonishing that they must represent technologies that are advanced beyond any available on Earth. Clearly, these sightings point to the involvement of space aliens—or perhaps just a global cabal of nefarious humans with ultraspiffy, above-top-secret flying machines that routinely break the known laws of physics. At least, that’s what modern-day folklore would have you believe, no matter how many times skeptics convincingly debunk sensational UFO sightings as mere misidentifications of conventional aircraft, sensor artifacts or natural phenomena. Regardless of what one personally believes about all this, what’s certain is that claims of mysterious trespassers in American airspace are taken very seriously by the U.S. government for reasons of national security. That’s why, at Congress’s behest, the U.S. Department of Defense established its All-Domain Anomaly Resolution Office (AARO) in July 2022. This office investigates reports of UFOs under the more generic rebranding of unidentified anomalous phenomena (UAPs). AARO’s work, however, isn’t really about chasing down extraterrestrial invaders so much as it involves standardizing reporting methods, curating and analyzing datasets and assessing possible threats posed by UAPs. Think less Men in Black and more 'Pentagon desk jockeys with advanced degrees and highly classified résumés.' The office’s current director Jon Kosloski, who took over in August 2024, after the departure of his predecessor Sean Kirkpatrick in December 2023, is a good example of the archetype. His professional past is punctuated by National Security Agency research in networking and computing, optical light communications and cryptography, as well as his invention of an advanced language-agnostic search engine for the DOD. Kosloski spoke with Scientific American about his vision and agenda for AARO—and, of course, his thoughts on the 'extraterrestrial hypothesis' for unexplained UAPs. What milestones and directions do you and AARO foresee as you assess the UAP situation? Kosloski: We have quite a lot going on in the office. We’re working diligently to make our AARO website available to the public so they can report an incident. We want to have a semiautomated processing chain to quickly bring those cases in and then look for correlations with other cases from government sources. We’re also working on machine-learning and artificial intelligence tools so that we can look for correlations at a larger scale. AARO is also looking at better utilization of the whole fabric of U.S. government sensors that are available. That will help our case resolutions, but it also poses the potential challenge of looking at an awful lot of data. A last big push is for increased transparency—to find more efficient ways to share information with the public and the scientific community to help us in some of our investigations. Standardizing the UAP data is an initiative—to make it better suited for data science. As we do that, we’d like having both the raw data as well as postprocessing data available to as wide of an audience as possible. But we need to respect the sensitivities of the sources and methods used to gather those data. Last November you testified before the Senate Committee on Armed Services’ Subcommittee on Emerging Threats and Capabilities. One point you raised was the importance of removing the stigma of reporting a UAP event. How much of a problem is that? Kosloski: It seems to have diminished quite a bit, thankfully, through a variety of public endeavors. I think it does persist, for example, with local law enforcement and some military members, however. They have experienced some pushback themselves from discussing events that they’ve been a part of but have come forward and shared that information with AARO. So maybe the stigma is reduced but not completely gone. Good progress is being made, but probably there’s a little ways to go. AARO has reportedly deployed the Gremlin System—a multisensor networked system for detecting, tracking and characterizing UAPs—in an undisclosed location. What is the status of this project? Kosloski: Gremlin is envisioned to be a test bed for sensor evaluations and sensor fusion. We expect it to be a 'living' system, always evolving. We will be swapping in and out new sensors and algorithms for various approaches. We want to document what we learn and share that with the public and other organizations inside the government. The hope is they can take what we have learned and replicate that. Gremlin itself is going quite well and is very robust. It has been running out in the wild for several months now, gathering data using radar, electro-optical and infrared cameras and some electromagnetic sensors. Gremlin is pulling all that in and detecting events in real time. We haven’t found anything particularly interesting yet.... But there are a few interesting sightings that are worth investigation. What does 'a few interesting sightings' mean, exactly? Kosloski: Interesting from our perspective. It is not obvious what the object is. It is unidentified. It has some anomalous characteristics. And in these cases, the object appears to be [moving] rather quickly. But it doesn’t appear to [be using] standard aircraft beacons. There could be multiple explanations for that, so we’re not jumping to any conclusions. But they are worthy of further investigation. UFO groups continually demand 'full disclosure,' by which they mean more than transparent data sharing—namely, the full disclosure of putative secret evidence that Earth has been visited by some form of alien intelligence. For AARO, if you come up with a head-scratcher of an unknown phenomenon, how would you disclose that fact? Kosloski: There are two things we have to consider. One is the owner of the data. We would need to work with the data owner to make sure that we are not revealing sensitive information about sources and methods. But putting that aside, there is nothing inherently classified about an anomalous event or phenomenon. So we would work with our leadership to document that well, study it and then produce a product that can be shared widely with the community. There is no inherent reason why we would sit on these anomalies, if we did come across something truly perplexing. Given that many full-disclosure advocates would insist that the absence of evidence for alien visitation just further confirmed a government cover-up, it seems like any push you could make for transparency as leader of AARO would, in some respects, be doomed to failure. So why take on the assignment in the first place? Kosloski: I love difficult scientific challenges. In general, I enjoy the chase of trying to understand the mystery and solving puzzles. UAPs, by their very nature, are rare events. [That means] data acquisition is difficult, and there’s a sensor design challenge. I think getting the data we need has been somewhat neglected. There are groups, such as Avi Loeb’s Galileo Project, that are working on better data acquisition—which is fantastic. But we in the U.S. government could do better. As a mathematician and data scientist, I enjoy poring through the data, looking for the subtle correlations and teasing out the threads to identify hypotheses and get the scientific method started. Doing that with a well-qualified group—with experts from inside AARO, as well as from across the U.S. government, and hopefully partnering with academia, too—was just too good of an opportunity to pass up. How does AARO engage academia? Kosloski: Right now we are primarily focused on a few key partnerships with University Affiliated Research Centers [UARCs]. For example, AARO worked with the Georgia Tech Research Institute to develop the Gremlin System. As a UARC, the Georgia Tech Research Institute partnered with AARO. Because of the nature of the contractual relationships, we can share the data that we have. In broader engagements with academia, there is the key step of downgrading classification to ensure we can release the information. We are working on that. Also, an important partnership is with the National Laboratories—Oak Ridge National Laboratory, for instance, has assisted AARO with some of its material analysis. But right now the focus has been on those few key UARCs. Is it really accurate to say you’re being 'transparent' while also dealing with sensitive, classified data? Kosloski: It’s a challenge. AARO should be as transparent as possible, and we are working in that direction. At the same time, we need to respect the sensitivities of the sources and methods used to gather the information that is relevant to UAPs. Ultimately, we are also working with our partners to release as much information as possible about the full context in which that data was gathered. It is a time-consuming process. But there are reasons why the U.S. government needs to protect those sources and methods so that we don’t put them at risk. Recent reports of mysterious drones over New Jersey and elsewhere sparked a lot of public interest and discussion—and I’m sure you and your AARO colleagues were paying close attention, too. What’s your take on that situation? Kosloski: Misidentification does account for a number of UAP sightings. AARO has been working on educational materials about common misunderstandings, such as Starlink flaring, as we call it, or [confusing visual] phenomena such as parallax. We’re sharing that with the public so that they understand what they are looking at. Specifically, with the New Jersey incidents, none of those were reported to AARO as being anomalous. Certainly, many of them were unidentified, whether they were drones or airplanes or other objects. They weren’t behaving in a way that was identified by the people at the time as being anomalous, so we didn’t take the lead on any of those investigations. But we were in contact with a number of federal organizations, offering our support. Do you have any advice for true believers who are convinced that aliens are visiting Earth’s skies? Kosloski: I don’t have advice per se. I don’t want to be the thought police. I think everyone is welcome to approach this topic however they like. As I came into this subject, however, I tried to approach it without bias in either direction. I’m open to any possibility. Sometimes an unusual event is just a sensor artifact, sometimes just a balloon. We do have some events in our holdings that are really peculiar, and I don’t know yet what’s behind those. But because we don’t know what’s behind them, we also can’t attribute them to anything in particular. And that includes extraterrestrial sources. Any closing thoughts on your primary objectives for AARO? And how do you feel about the big, daunting question 'Are we alone?' Kosloski: I’m impatient and, being a data scientist, I am also a data hog. I want more data, and I want it quicker so that we can get to the heart of these problems. It boils down to asking ourselves, 'What type of data is it going to take to prove to the scientific community, beyond a shadow of a doubt, that this anomalous phenomenon was not a sensor artifact and is, indeed, something truly peculiar?' And I feel that’s going to require multiple sensors gathering the same event from different perspectives at the same time. My goal is for AARO to be able to enhance our national security by increasing domain awareness, ensuring that we understand everything that is operating in space, in the air and maritime environments, as well as those trans-medium objects [UAPs that seemingly slip between, say, the sky and the sea]. I think it’s plausible that there’s life out there. I haven’t yet seen the substantial evidence I need to convince me that extraterrestrial life has found its way to Earth as yet, but I am open to anything.