Let’s use AI to stop fentanyl at the border and keep it from killing Americans
Over 70,000 Americans were killed by fentanyl. AI is an inexpensive solution to help us stop drugs at the border
Originally published April 18, 2024 on Fox News Media. You can read the full opinion piece here.
Over 112,000 Americans overdosed in 2023. Over 70,000 overdosed on fentanyl, a particularly dangerous synthetic opioid. While it’s easy to focus on numbers, we can’t forget that these aren’t statistics. They’re people. They leave behind parents, brothers, sisters and children. They’re futures lost. They’re lives squandered.
We owe it to victims of this epidemic to do everything we can to curb the flow of fentanyl from overseas (where the vast majority of it originates) into American cities and towns. This poison is a direct threat to the security of America’s citizens. And, though America’s borders are a point of grave concern, they are also the best opportunity we have to seize fentanyl and damage the cartels responsible.
And here, there is good news. The Department of Homeland Security (DHS) and the Customs and Border Protection Agency (CBP) can leverage the power of artificial intelligence (AI) to identify the trucks, boats and planes trying to sneak fentanyl into the country.
We must use this technology at the border and ports of entry (where nearly 85% of America’s fentanyl comes into the country,) when we have access to every vehicle coming in.
AI (Artificial Intelligence) represents an inexpensive and advanced way to search for illegal drugs at the border with minimal manpower. (REUTERS/Dado Ruvic/Illustration/File Photo)
We may never completely stop the flow of fentanyl, but we can curb it. We can make sure less of it gets to our streets and make it less profitable for the cartels behind it. We can raise its price and keep it out of reach for more Americans.
While AI is complex, the concept behind using AI to stop fentanyl is straightforward. AI enables pattern recognition on an enormous scale and "translation" of data into simpler interfaces for human domain experts. Gemini, ChatGPT and Claude can write responses to prompts by looking at tens of billions of pieces of online content, seeing what people have said, and repeating the most common patterns.
AI at our ports of entry would do the same thing, albeit in a more targeted manner. It would, for instance, examine the characteristics of every truck that has ever been caught smuggling fentanyl. Where had they stopped beforehand? How far into America had they stopped once getting past the border on previous trips? What cargos had they claimed to be hauling?
That’s only counting the patterns that seem relevant to drug smuggling, but there are countless other patterns that would emerge that human analysts may never otherwise consider. For all we know, there could be a correlation between propensity to smuggle drugs and a brand of tire or a truck’s color.
AI could examine all this data and tell us exactly which vehicles are likely smuggling fentanyl. Then, instead of conducting random searches – which find fentanyl in just a handful of trucks of the over 19,000 that cross the southern border every day – CBP could conduct targeted, data-informed searches. An AI-based system would flag to CBP officers the trucks they needed to search. The "hit rate" would be far higher than it is today.
These drug seizures, while valuable on their own, don’t even count the second order effects. It would make smuggling fentanyl less profitable; cartels outlay enormous sums to manufacture fentanyl and get it across the border.
When we seize it, that investment is lost without a payoff. Years of losses might then convince some cartels that fentanyl smuggling just isn’t worth it, further limiting the fentanyl supply. After all, cartels care about profit, not ideology. They will abandon efforts that don’t make them money.
Further, cutting cartel profits bolsters our national security. Cartels use those profits on powerful weaponry and de facto soldiers, which they use to control large swathes of Mexico. Such regions, often inaccessible to U.S. or Mexican law enforcement, are dangerous to have near our border; we do not know what cartels would, for the right price, allow America’s adversaries to cook up in those areas.
Closer to home, AI would allow CBP to make better use of its most important yet overstretched asset, its officers. Examining trucks is time consuming. Every minute not spent checking harmless cargo can be spent stopping other smuggling operations. Given that human trafficking is a major problem, it’s crucial agents have more available time and are where they have maximum impact.
Further, while CBP does data-based analysis, those efforts can take days. When AI can do the same tasks in minutes, it frees up dozens of agent hours.
Finally, AI can help track fentanyl to its source. Given enough data, we will be able to detect if fentanyl is coming from Mexican factories, Chinese manufacturers, or some other source we hadn’t discovered. And, once we know a source, we can stamp it out.
These drug seizures, while valuable on their own, don’t even count the second order effects. It would make smuggling fentanyl less profitable; cartels outlay enormous sums to manufacture fentanyl and get it across the border.
With AI comes another critical benefit, particularly when Congress is failing to appropriate the funds America needs to secure its border: it is inexpensive. An AI program targeted at fentanyl smuggling would cost less than $10 million. It is cheaper than border surveillance drones, which cost $17 million up front and over $12,000 an hour to operate.
AI also looks like a steal compared to the $30-billion price tag for the "virtual wall." That’s not to say DHS and CBP shouldn’t have access to the tools needed to secure the border, but these agencies should prioritize tools that have the highest return on investment.
The fentanyl crisis cannot wait. Political maneuvering and bickering are problematic, but when they halt tangible solutions that could save thousands of Americans, they are downright immoral. When just a few million dollars can save so many lives, it would be unforgivable for Congress and the Biden administration not to act.