ESEN
Contact us
← Back to the blog
Market GABOTRIX AI 5 min read

16-year-old founds AI missile startup: what it means

According to Wired, a 16-year-old has founded an AI missile startup focused on "cost per kill." What it reveals about the AI market, and why it matters to businesses in Colombia.

ShareWhatsAppX
Article illustration: 16-year-old founds AI missile startup: what it means

According to Wired, a 16-year-old has founded an artificial intelligence missile startup with the goal of reducing "cost per kill", that is, the cost per casualty or per neutralized target. The story brings together two facts that, taken together, say a great deal about where AI stands today: a teenager can set up a military technology company and describe his product in the language of efficiency metrics, like any other startup. Beyond the headline, it is still unclear what stage the product is at, who is funding it or how it works.

Why an AI missile startup is news for the business world

At first glance, this may seem like a distant story for a small business in Colombia. It isn't. What the case shows is that the barriers to building AI systems have fallen dramatically. Today there are computer vision models, open-source tools, cheap hardware and assistants that write software. That allows very small teams, and even people with no track record, to build products that a decade ago would have required entire laboratories.

The same democratization explains why a logistics company in Medellín can have an agent that handles orders over WhatsApp, or why a clinic in the Eje Cafetero can automate its scheduling. The technology is the same. What changes is the purpose, and that is where the responsibility of those who build it, and those who buy it, comes in.

At the same time, defense technology has gained ground in conversations among investors and founders in recent years. The fact that an outlet like Wired is devoting coverage to a teenage founder in this sector signals where part of the market's interest is heading.

"Cost per kill": when the metric says more than the product

The most striking phrase in the headline is not the founder's age but the metric. Reducing "cost per kill" means applying the logic of unit economics to a lethal outcome. It is the same reasoning a company uses for customer acquisition cost or cost per resolved ticket, transferred to a domain where the "outcome" is a human life.

For any company, the lesson is uncomfortable but useful: metrics shape the behavior of both systems and teams. If an AI agent is optimized only to close sales, it may pressure customers. If it is optimized only to cut handling time, it may cut short conversations that needed more care. Deciding what to measure is an ethical and a business decision at the same time, not a technical detail.

Dual-use AI: a risk that also affects Colombian companies

The key concept here is dual use: the same capability that recognizes products in a warehouse can recognize targets; the same autonomous navigation that guides an agricultural drone can guide an armed one. Colombia has first-hand experience of drones being used in armed conflict, so this discussion is not theoretical.

For a Colombian company adopting AI, dual use translates into concrete questions:

  • What does my provider do with my data? Customer conversations, images and records could end up feeding models for purposes the company would never approve.
  • What usage policies apply to the models I contract? Several major labs have revised their rules on military and security applications in recent years. It pays to read them rather than assume.
  • Who am I sharing a supply chain with? Reputational risk is real: a client or international partner may ask what technologies your company uses and where they come from.

What is changing in the market and in choosing providers

The case reflects a market where anyone can build and sell AI, with very little friction. That makes the technology cheaper and speeds up innovation, but it also fills the market with providers that lack data governance, traceability and clear criteria on what they will and will not do with their technology.

For companies in Colombia, the practical consequence is that choosing an AI provider is no longer just a matter of price or features. It is also a matter of trust: where data is processed, who has access, what usage limits the provider has and whether it can explain them clearly. Anyone working in artificial intelligence for business should be able to answer those questions directly.

Regulation is advancing, but more slowly than the technology. In Colombia, the debate over rules for AI remains open, and until it is settled, responsibility falls largely on companies themselves and their contracts.

What a company in Colombia can do today

  • Review the contracts and terms of use for every AI tool: data handling, model training and prohibited uses.
  • Set an internal AI policy that specifies which tasks are automated, with what data and under what human oversight.
  • Choose balanced metrics: not just efficiency, but also customer satisfaction, response quality and cases escalated to a person.
  • Ask providers about traceability: the ability to audit what an agent did and why.
  • Comply with Ley 1581 de 2012, Colombia's personal data protection law, in every workflow that uses AI.

At GABOTRIX, we work with agents that run business processes such as customer service, sales and systems integration, and we believe the conversation about what AI is used for has to be on the table from day one, not when a problem arises.

Frequently asked questions

What is dual-use AI?

It is artificial intelligence technology that can serve both civilian and military purposes. A vision model that identifies products in an inventory can, with other adjustments, identify targets. That is why serious providers set clear policies on which uses they allow and which they do not.

Does an AI missile startup affect a Colombian small business?

Not directly, but it does show how the market works today: building AI is easier than ever, and there are providers of every kind. For a small business, the consequence is that it needs to vet more carefully who it trusts with its data and processes.

How do I choose a trustworthy AI provider in Colombia?

Ask where your data is processed, whether it is used to train models, what human oversight exists and how the system's actions are audited. A trustworthy provider answers in writing and without ambiguity. Vague answers are a red flag.

If your company is evaluating how to adopt AI with clear rules on data and oversight, let's talk about your case.