A Latin American AI revolution?
Hello and welcome to the latest edition of The Counterbalance. This week, we’re taking a look at the fight back against Big Tech in Latin America.
In Latin America today, a quiet revolution is taking shape — one that could redefine how we think about artificial intelligence, cultural sovereignty, and the unchecked power of Big Tech.
Its name? Latam-GPT.
Latam-GPT is a homegrown, community-driven language model that is being designed specifically for the diverse cultures, dialects, and social realities of Latin America. It’s an AI model that speaks the region's languages, literally and figuratively.
Set to launch in September 2025, Latam GPT’s emergence is being championed as an act of resistance against the all-encompassing power and influence of Big Tech.
For years, Silicon Valley’s tech giants have promised to connect the world. But what they didn’t say is that, in doing so, they would standardise it too. The same AI models that are being created in the United States are being dropped — wholesale – into São Paolo, Bogota, and Buenos Aires.
When a handful of US-based companies control the tools that promise to revolutionise how we communicate, they don’t just dominate markets, they shape narratives, identities and entire cultures. The result is a flattening of local nuance, the erasure of global slang, and a steady erosion of cultural identity in favour of established norms from the Global North.
The story of Latam-GPT is important because it reminds those involved in the anti-monopoly fight to broaden their horizons. Society typically thinks about monopolies for their economic harms — price gouging, unfair competition, worker exploitation, and stealing copyrighted material — but Big Tech’s stranglehold over AI means the social harms are just as urgent and impactful.
“We talked about creating technology in this part of the world, technology that thinks, acts, and decides like us. How many times have we wondered why smartphones are not manufactured in Chile, why cars are not built in Peru?” said Omar Florez, a technical lead involved in training the Latam-GPT model.
Loyal subscribers to The Counterbalance will know there is precedent for tech giants sweeping cultural sensitivities aside in the pursuit of profit and market dominance.
A 2024 study from The Netherlands found digital streaming platforms like Spotify having a “generic popularity bias” for content that exposed the need to “improve representation of users with non-mainstream music preferences.”
According to the same study, the inclusion of AI in music generation contributed to a “critical gap in the fair representation and inclusion of the music genres of the Global South.”
Latam-GPT is the first Large Language Model being developed collaboratively in Latin America — and the Caribbean — and being coordinated by Chile’s National Center for Artificial Intelligence (CENIA). The initiative is a signal that the Global South won’t idly stand by while Big Tech steamrolls over our cultural landscapes.
“Language models that have been generated in the Global North…do not necessarily reflect the idiosyncrasy and nuances of Latin America and the Caribbean,” the CENIA website reads. “It is important that in the region we can develop capacities to have a certain independence and make decisions about how this technology impacts society,” adds CENIA director Alvaro Soto.
Of course, Big Tech has nearly limitless resources. By the very nature of monopoly power, these tech giants can outspend, out-market and out-lobby local innovators. Today, two of the most popular LLMs — Chat GPT by OpenAI and Gemini by Google — have over 300 million weekly active users.
But crucially, the role of Latam-GPT is not to break Big Tech apart, but to provide a vital cultural and social counterweight, reminding us that technology does not have to be a foreign export but, instead, can be rooted in local purpose.
“Latam-GPT won’t compete with global tech giants: instead, it will power tools for education, healthcare, and public services,” said Kostas Rossoglou, Director of International Public Policy and Government Affairs at Shopify. “As Latin America brings its voice to the global AI conversation, Latam-GPT is one to watch.”
Weekly highlights:
The UK government this week unveiled a new Financial Services Growth and Competitiveness Strategy which, in their words, “focuses on delivering a competitive regulatory environment.” This includes shorter applications for businesses seeking to gain FCA licenses, as well as “focusing on priority growth opportunities” that make the UK an attractive place for investment. The government’s new approach is another example of how deeply embedded the narrative of “deregulation = growth’ has become in UK political circles. As pointed out by Jesse Griffiths, CEO of The Finance Innovation Lab, a push for deregulation “risks a race to the bottom.” “Without confronting the structural reasons why the financial system has failed to support the real economy, while risking a race to the bottom in regulation, the strategy risks repeating old mistakes,” Griffiths added.
The European Commission is considering a contentious proposed acquisition of Downtown by UMG that indie labels fear could devastate their sector, and that The Counterbalance covered earlier this year. Via POLITCO, Martin Mills — a record producer and founder of Beggars Group, the label that signed Adele before her worldwide fame — said “if it were not for the independents, there would be no hip hop, no punk, and no Elvis Presley.”
Soundbite of the week: Impassioned advocacy by former Meta executive
Meta’s former head of public policy Aura Salla took to LinkedIn earlier this week to call on Europe to issue a strong response to US tariffs.
In her post — which you can read in full here — Salla claimed that the EU can “live without the products of Meta and X” and that Europe must take tougher action to secure its technological and digital sovereignty.
“I would not exclude services and data from the negotiations either. We must consider how to put pressure on the US. This includes, among others, Big Tech. Retaliatory tariffs against US social media platforms should not be excluded from our toolbox. All options must be on the table if we are to defend our interests.”
Data mining: AI mega-deals drive surge in venture capital dealmaking
Fresh data from analytics platform PitchBook reveals that almost $84 billion of private capital was deployed across almost 500 deals in the venture capital sector in the first half of this year.
On an annualised basis, this suggests 2025 could reach highs of $167.8 billion invested in growth-stage venture capital deployment, which would far surpass the current year-long high of $91.6 billion in 2021.
These figures highlight a mounting risk that immense pools of private capital are reinforcing the dominance of already-powerful players, and concentrating power in fewer hands. The sheer size of these deals means that emerging technologies can be locked up by well-funded giants.
AI continues to dominate the upper end of the deal spectrum, notably via OpenAI closing a $40 billion funding round in March of this year, which amounted to the largest private tech funding round in recorded financial history.
The Counterbalance is published every Thursday. Please send any thoughts and feedback to scott@balancedeconomy.org.