Military artificial intelligence has a new testing ground: US troops in Europe. They are getting ready for a kind of attack that is hard to track with human eyes alone, with missiles and drones arriving in swarms and a huge stack of targets to analyze all at once. According to a Business Insider report published on Yahoo, the benchmark the military is using is what it sees as the worst day Ukraine has faced in the war.
The plan, though, is not to swap soldiers for machines. The technology steps in to filter out the overload, sort through the mess and point to possible paths. The final call stays with the person who knows the subject and answers for what happens next.
And this is where the story reaches you. Your company probably does not deal with drones. But it may get a flood of tickets, alerts, leads and reports every day. The logic is similar: let the machine pick out what matters and save the minds of your team for the cases that call for real judgment. Let us take it step by step.
What US forces are testing in Europe
According to the Business Insider report, the American command on the continent is working with a heavy estimate. The math points to more than 1,500 potential targets per day, plus up to 1,200 missiles and drones heading toward the troops. That is far too much for any room full of analysts, no matter how experienced they are.
These numbers were presented by Brigadier General Steven Carpenter, head of Multi-Domain Command Europe. According to the outlet, he explained that the estimates reflect what the military believes was the worst day for Ukraine. The comparison makes sense. The war of Russia against the country has been marked by frequent, heavy bombardments that mix ballistic missiles, cruise missiles and attack drones. The report points out that some of these drones now come with jet engines, which makes them faster and harder to shoot down.
The topic also came up at the Association of the United States Army convention in Washington last month. There, according to the outlet, leaders and soldiers talked about using AI to process battlefield data and support decisions without taking people out of the process.
Hands-on testing is already underway. The report says that this year, a joint exercise between the US Army and NATO used an AI system built by the Army itself to take in and organize data. It was still in beta, and the goal is to test it on a larger scale in a future exercise. At the same time, the Army has been adding AI to NGC2, its new command and control system, which pulls together real-time information on targets, capabilities, ammunition, maintenance, intelligence and even simulated enemy attacks. If you want to see another chapter of this same debate, it is worth reading about target oversight with Palantir in the military use of AI.
Receive the best innovation content in your email.
All the news, tips, trends, and resources you're looking for, delivered to your inbox.
By subscribing to the newsletter, you agree to receive communications from Método Viral. We are committed to always protecting and respecting your privacy.
AI for threat triage: speed and scale where humans get stuck
The way it works, as described in the report, is simpler than it sounds. AI for threat triage takes in data from sensors and networks, cross-checks that information and ranks targets based on the goals set by the commander. Then it reviews the available resources and suggests responses. Those suggestions can range from a strike with conventional weapons to an electronic warfare or cyber action.
Carpenter summed up the thinking well. Speaking to Business Insider, he said the technology should be used for what people are not so good at, which is speed and scale. In the same remarks, he made it clear that he does not think humans should be taken fully out of the loop.
Speed is the most obvious gain, but it is not the only one. According to the outlet, the machine can also remember and spot patterns that a person might forget or never even notice. Picture someone after hours of staring at screens, under pressure. Fatigue takes its toll, and details slip by. Software does not get tired in the same way, though it has its own flaws, as we will see later on.
The most interesting point, though, is the stated goal. Carpenter said the idea is not to replace analysts but to take weight off their shoulders so they can focus on the truly hard targets. The example he gave was dual-use facilities, which can serve both civilian and military purposes and require much more careful analysis.
The machine clears the ground. The human handles the delicate part.
Automation with human oversight is not a detail, it is the model
In the words of the general, the soldier still has the final say. According to the report, it is the person who decides whether a target is valid, weighing everything that must be respected under the laws of armed conflict. AI offers options. It does not act on its own.
In the business world, this logic has a name: automation with human oversight. The system recommends, organizes and prioritizes. Someone who is accountable for the outcome reviews and decides. It may look like an extra step, almost red tape. In practice, it is what lets you use the technology without losing control of what is going on.
Taking humans out of the loop on sensitive decisions is risky for several reasons. AI models make mistakes, sometimes with a lot of confidence. Input data can be incomplete, outdated or biased. And when something goes wrong, someone needs to be able to explain why that choice was made. Without a person in the middle, the company has no answer for the customer, for its own team and, in some cases, for the courts.
Imagine, just as an example, a lender that let a system approve and deny credit without any review. A single badly entered piece of data could be enough for good customers to be turned down one after another, and nobody would be able to point to the reason clearly. With a person reviewing the unusual cases, the error tends to show up before it turns into a crisis.
AI in decision-making: the parallel with everyday business
Nobody here is saying that a call center looks like a battlefield. The stakes are on a whole different level. But the core problem is similar: lots of things coming in at once and too few people to look at everything calmly. That is where AI in decision-making starts to make sense for your business.
A few common scenarios help paint the picture:
- A support team that gets lots of tickets a day and needs to separate the customer whose system is down from the one who just wants to reset a password.
- A cybersecurity team dealing with an avalanche of possible fraud alerts, many of them false.
- A marketing team tracking mentions, leads and campaigns at high volume that cannot figure out, on its own, what deserves an immediate reply.
In all of these cases, the machine can do the first sweep and hand over a queue sorted by priority. Picture an online store during a big sale. The AI flags orders with odd signals, and an analyst reviews only those instead of opening every single one. The analyst does not disappear. They start spending energy where it makes a difference.
Another illustrative example: an agency that runs social media for several clients. Instead of someone reading every comment of the day, a tool groups messages by topic and highlights serious complaints. The team responds first to what could become a reputation problem and leaves the compliments for later. To dig deeper into this line of thinking, we have a full piece on artificial intelligence in decision-making.
Automation for Brazilian businesses: how to start with your feet on the ground
The first step in automation for Brazilian businesses is not choosing a tool. It is finding out where volume is smothering your team. It could be the support inbox, the lead spreadsheet or the alert dashboard nobody can keep up with anymore. Start with the bottleneck that bothers you most.
Next, spell out clearly what AI can handle on its own and what needs to go through a person. Answering a simple question about delivery times? That can be automatic. Canceling an account, denying a refund or blocking a customer over suspected fraud? That is where human approval pays off. Put these rules in writing, track the errors and only expand the use once you are comfortable with the results. If you want to see how this fits into daily operations, check out our piece on AI and automation in workflows.
Notice that not even the US Army rolled everything out at once. The system used in the NATO exercise, according to the report, was still in beta, with a bigger test planned down the road. For a company, the message is the same. Starting small, in a controlled process with people keeping a close eye on it, is usually the safest route. It gives you time to adjust course before a mistake turns into a loss.
A practical tip: pick a simple metric to tell whether the change is helping. It could be the time to first response for customers or the number of cases the team can handle without rushing. Without some kind of measurement, it is hard to tell real gains apart from a mere feeling of being modern.
People thinking better, machines handling the overload
The lesson coming out of Europe is almost counterintuitive. One of the sectors that invests the most in technology is treating AI as a tool to give time and attention back to people, not to erase them from the process. The machine runs. The person weighs.
For anyone running a business or working in marketing or tech in Brazil, this changes the question. Instead of asking how much of the team can be replaced by AI, the question becomes: which decisions deserve more of the attention of the team, and how can AI make room for them?
This balance does not happen by accident. It depends on clear rules, honest testing and someone who takes responsibility for the final outcome. Companies that get this mix right tend to gain agility without giving up the trust of customers and employees.
A good exercise is to pick one process in your operation where volume runs over the team and sketch out, on paper, what the machine could filter and what would stay in the hands of a person. To keep going down this path, keep exploring the Método Viral content on AI and automation for business, like the article on AI and automation for businesses.
