October 1984 smelled of rain and hot coffee in Brazil. In Rio de Janeiro, tourists posed for photos against the backdrop of Christ the Redeemer, unaware that 2,500 kilometers to the south, in the capital Brasília, history was being made—history that would upend the global tech industry. On October 29, President João Figueiredo, the last general of the military dictatorship era, signed a document that turned the country into a laboratory of digital isolationism. Law No. 7.232, known as the "Informatics Law," didn’t just regulate the market—it declared war. A war against IBM mainframes, against Intel chips, against software from across the ocean. A war in which Brazil decided to fight alone, armed only with pride, import substitution, and the belief that technology could be grown like coffee.
By the early 1980s, Brazil was a country of paradoxes. A country where ATMs already stood on the streets of São Paulo, while government offices still typed documents on Olivetti typewriters. A country where university students wrote programs on punch cards, and local companies dreamed of owning their own computer—but ran into reality: 90% of the IT market was controlled by foreign corporations. IBM sold System/370 mainframes, Burroughs supplied machines for banks, DEC provided minicomputers for universities. Brazilian companies, if they appeared at all, existed only as middlemen, reselling foreign hardware and software.
The problem wasn’t just money—though that was part of it. Every imported computer meant capital flight, and Brazil, drowning in debt after the 1973 oil crisis, couldn’t afford it. But there was another, deeper reason. The military, which had come to power in 1964, saw technology not just as business, but as a matter of national security. If all of a country’s computers depended on the U.S., then Washington could cut off the oxygen at any moment. What if Brazil decided to pursue an independent foreign policy tomorrow? What if it started trading with Cuba or supported the Sandinistas in Nicaragua? The CIA had already shown what it was capable of when, in 1975, it blocked computer shipments to Chile after Allende’s rise to power. The Brazilian generals didn’t want to end up in the same trap.
In 1979, the government created the Special Secretariat for Informatics (SEI)—an agency tasked with controlling technology imports. But SEI was only the first step. The real battle unfolded later, when it became clear: bans alone wouldn’t change the market. Something bigger was needed. Brazil had to be convinced it could produce what it bought from others.
Law No. 7.232 wasn’t just a set of rules—it was a manifesto. Eight pages of text, signed by Figueiredo and Minister Danilo Venturini, turned Brazil into a fortress with a drawbridge. Article 4 of the law imposed an eight-year moratorium on imports of any informatics goods and services if domestic equivalents could be produced. This applied to everything: from personal computers to operating systems, from microprocessors to banking software. Exceptions were made only for what Brazil couldn’t produce itself—but even then, imports were hit with customs duties of up to 100%.
But the law went further. It created the National Council for Informatics and Automation (CONIN)—a body that decided which companies could operate in the market and which couldn’t. To earn "national" status, a firm had to meet strict criteria: 51% of shares had to be Brazilian-owned, technology had to be developed domestically, and key decisions had to be made without foreign influence. Foreign giants like IBM could stay, but only as minority partners or producers of outdated equipment.
The main tool of the new policy was the Fund for the Development of Informatics (FNDCT). Money flowed into it from taxes on imports and sales of foreign equipment, and was spent on grants, subsidies, and the creation of research centers. One of them, the Center for Information Technology (CTI), was meant to become Brazil’s answer to Bell Labs—a place where its own processors, operating systems, and computers would be born.
Brazil decided to build digital sovereignty on the bones of the free market. But the market doesn’t like being told what to do.
The first child of the new policy was the G-10—a personal computer developed by the Brazilian company Microdigital. Outwardly, it resembled the Apple II, but inside, it was one big compromise. The Z80 processor (a clone of the Intel 8080) ran at 4 MHz, RAM was 64 KB, and the CP/M operating system was archaic even by early 1980s standards. But the main problem wasn’t the specs. The G-10 cost $3,000—twice as much as the IBM PC, which was already conquering the world by then.
Brazilian companies, forbidden by law from buying foreign computers, were forced to take the G-10—not because it was good, but because there was no choice. The government subsidized production, but demand remained sluggish. In 1985, a year after the law was passed, Brazil produced 12,000 personal computers—while the U.S. sold 2 million. Even in neighboring Argentina, where there were no bans, computers were cheaper and faster.
The situation was worsened by the fact that Brazilian companies couldn’t compete even in their own market. IBM kept selling mainframes because there was nothing to replace them, while Apple and Commodore simply ignored the bans, supplying computers through third countries. Smuggling thrived: in 1986, customs seized 5,000 illegal computers, but that was just the tip of the iceberg. Experts estimated that the black market had twice as many devices as were produced legally.
But the biggest failure was the attempt to create a homegrown operating system. SISNE—Brazil’s clone of MS-DOS—launched in 1987, but was so unstable that even government agencies refused to use it. Programmers joked that SISNE stood for "Sistema Instável, Sem Ninguém Esperar" ("Unstable System, Nobody Was Waiting for It"). In the end, most Brazilian companies just pirated MS-DOS, despite the bans.
By 1990, it was clear: the experiment had failed. Brazilian IT was a generation behind the rest of the world. While the U.S. was already selling IBM PS/2 machines with Intel 80386 processors, and Japan’s NEC was releasing its first laptops, Brazilian companies were still assembling computers based on the Z80. The Informatics Law hadn’t just failed to create a competitive industry—it had isolated the country from technological progress.
But the most paradoxical part was that the law wasn’t repealed even after the fall of the military dictatorship. In 1985, the civilian president José Sarney came to power, but he didn’t dare touch SEI or CONIN. Only in 1991, under pressure from the U.S. and the World Trade Organization, did Brazil begin to gradually ease the bans. By 1992, computer imports were allowed again, and by 2000, the old policy was just a memory.
Yet there’s another side to this story. Brazilian companies that grew up under the law’s protection learned to survive in isolation. Positivo, a computer manufacturer that started by assembling the G-10, is now the largest IT vendor in Latin America. Totvs, a developer of corporate software, became a regional leader because in the 1980s, it was forced to build software from scratch. Even CTI, despite all its failures, survived and today works on microelectronics development.
Brazil didn’t manage to build digital sovereignty. But it proved that even in the age of globalization, you can try to go your own way—even at the cost of falling behind. Today, as the world talks again about technological independence, sanctions, and bans, the story of Law No. 7.232 sounds like a warning. Sovereignty isn’t about walls built around the market. It’s about the ability to create what you can’t buy for any amount of money. Brazil didn’t do that. But at least it tried.