画像クレジット: Image via Dronelife. Used under fair use for news commentary. · source
The Department of Homeland Security says the United States is still building its counter-drone defenses for the 2026 World Cup, which kicks off in about twelve months. DHS Secretary Kristi Noem's office used the phrase "a little behind" to describe the current state of preparedness. I've been covering tech long enough to know what "a little behind" means in government speak: it means we're scrambling.
This is the self-driving car hype cycle all over again, except instead of overpromising on autonomy, we've been overpromising on security while the actual regulatory framework sits in committee somewhere gathering dust. The FAA, FBI, and local police departments are all pointing at each other wondering who's supposed to shoot down a rogue quadcopter when it buzzes a stadium full of 80,000 people. The answer, as of right now, appears to be: nobody knows!
Here's the thing that kills me about this whole situation. We've known drones were going to be a security headache since, what, 2015? Maybe earlier. I remember writing about drone incursions at airports back when the Phantom 3 was the hot new toy. That was over a decade ago. And yet here we are, with the biggest sporting event in American history coming to eleven cities across the country, and according to Dronelife, federal officials are still figuring out the basics.
The core issue isn't technical, it's legal. Local police can't legally jam drone signals. The FBI has limited authority to act. The FAA can designate no-fly zones but can't enforce them with anything more threatening than a sternly worded violation notice. Meanwhile, any kid with $500 and a YouTube tutorial can fly a drone wherever they want, and the worst case scenario (I'm told we're not supposed to spell it out, but you can imagine) keeps security professionals up at night.
関連記事
More in Drones
After years of lab demos that fell apart in real buildings, researchers are figuring out how to make drones and robots actually navigate using natural language commands.
Robert "Bob" Macintosh · 4 hours ago · 4 min
New research shows that how a model handles variability in training actually predicts whether it'll work on a real robot. Who knew?
Robert "Bob" Macintosh · 6 hours ago · 4 min
New geometric adaptive control research shows quadrotors can learn to fight wind disturbances in real-time. The theory's solid. The gap to industrial deployment? That's another story.
Robert "Bob" Macintosh · 18 hours ago · 3 min
The FBI and Atlanta PD recently held what appears to be a public awareness campaign about "no-drone zone compliance" for the World Cup matches scheduled in that city. Which is great, really, I'm sure the hobbyists and wedding photographers will comply. But the people we're actually worried about? They're not reading FAA advisories.
Call me old-fashioned, but I remember when the response to new technology was to figure out the rules before the crisis, not after. The internet had its share of growing pains, sure, but we at least had some framework in place before everyone and their grandmother got online. With drones, we've done the opposite. We let the consumer market explode, let the technology get cheap and capable and everywhere, and then started asking "wait, how do we regulate this?"
The counter-UAS industry has been begging for clearer authority for years. There are companies that make detection systems, jamming equipment, net guns, trained eagles (I'm not making that up), all sorts of solutions. But most of them can't be legally deployed by anyone except federal agencies, and even then the rules are murky. So we've got this absurd situation where the technology to stop drones exists, the threat is well understood, and we're still "a little behind" because nobody in Washington could agree on who gets to push the button.
The young founders in the counter-drone space, and I've talked to a few of them, they're pulling their hair out. They've got products ready to go. They've got customers lined up. But the regulatory environment is so uncertain that major deployments keep getting delayed or scaled back. One CEO told me (this was a couple years ago, via email, because that's how I prefer to communicate) that selling counter-drone tech in the U.S. was harder than selling it in the Middle East. Let that sink in.
The honest answer is we don't know yet. The World Cup will have matches in Atlanta, Dallas, Houston, Kansas City, Los Angeles, Miami, New York/New Jersey, Philadelphia, San Francisco, Seattle, and somewhere in Mexico and Canada too. That's a lot of stadiums, a lot of airspace, a lot of coordination between agencies that don't always talk to each other.
DHS says they're working on it. The FBI says they're coordinating. Local police departments are running awareness campaigns. But the fundamental authority question, who can actually do something when a drone appears where it shouldn't, that remains unclear. There's been some movement on Capitol Hill to expand counter-UAS authorities, but legislation moves slowly and the World Cup doesn't care about congressional timelines.
I'm not saying there's going to be an incident. I hope there isn't! The security professionals working on this are competent people doing their best within a broken system. But the fact that we're twelve months out from the biggest sporting event this country has hosted in decades, and federal officials are publicly admitting they're "a little behind" on drone defenses, well. It doesn't inspire confidence.
Maybe everything will be fine. Maybe the no-drone zone compliance campaigns will work perfectly and every hobbyist will check the B4UFLY app and stay home. Maybe the bad actors will decide the World Cup isn't worth the trouble. Maybe the patchwork of federal, state, and local authorities will somehow coordinate flawlessly despite having no clear chain of command.
But what do I know. I've just been watching this space for ten years, waiting for someone to figure out the rules. We're still waiting.
If you want to argue about this, my email's on the about page. I check it more than Slack, which my editor keeps reminding me is not how things work anymore.
Three new papers show reinforcement learning for drones is getting scary good at transferring from simulation to the real world. I've seen this inflection point before.
