TechinData.in Connect
AI & Geopolitics 101
Let's be honest — the mix of geopolitics and technology is cinema. Peak cinema. But not in the sense of spectacle or fiction. It May seem dense, as if you need jargons. What if we say it is not the case?
What is geopolitics for LLMs or any AI?
Here's the thing: engineers speak in models and datasets. Diplomats speak in treaties and strategic interests. When they're in the same room, they're often speaking past each other — one worried about algorithmic bias, the other about algorithmic hegemony. Same problem, different vocabulary.
And yet AI doesn't develop in a vacuum. Every algorithm trained reflects a technical worldview. It does not need to ultimately have a socio-political view.
Yet, market desperation, political posturing, marketing tactics, and manipulation of intellectual property laws create policy friction.
At least some of the above mentioned reasons of friction cause the geoeconomic dead-end. It's not entirely political, but it gets there.
In short, the tech and geopolitics bubbles speak their own languages, and patterns, making 0 sense.
Individualistic Sovereignty
Imagine you write a letter. Someone else reads it, makes copies, sells information about what you wrote, and you have zero say in any of this. Digital Sovereignty is fundamentally about YOUR right to control your own data, your own digital identity, and your own choices online. Of course, when some legal rights are limited, you see country-wise deviations across.
-
How It Works:
-
Every time you use an app, website, or AI tool, you generate data. Digital sovereignty means you—the individual—should have the power to decide what happens to that data.
Can companies surveil you?
Can foreign governments access it?
Can it be sold without your consent?
Should you trust in national courts to handle your grievances or another glorified North American senators briefing on tech companies? Ask yourself.
-
Normative Emergence
Imagine a few neighbours start composting in their backyards. Others notice, copy them. Soon it's the neighborhood norm. Eventually, the city makes it an official rule. Normative Emergence is when technical practices or informal behaviors gradually become accepted norms, and then sometimes become formal rules.
How It Works:
-
In 1994, a Netscape engineer invented cookies as a simple technical hack—just a way to remember items in a shopping cart between page loads. It was purely practical. No policy discussion. No debate. Just code.
-
Other developers saw it, copied it, and started using cookies for their own sites. Within a few years, advertisers discovered they could use "third-party cookies" to track users across multiple websites.
-
By the early 2000s, cookie-based tracking became the invisible foundation of online advertising. Every ad network, every recommendation system, every personalization engine assumed cookies existed and that tracking users across sites was just "how things work".
Normative Evasion
Imagine your local store sells plastic bags, but the neighboring town bans them. The store simply sets up shop a few meters across the border and keeps selling. Regulatory Arbitrage is when tech companies exploit differences in national laws to continue the same practices under friendlier jurisdictions.
How It Works:
-
AI companies locate their data centers, R&D hubs, or headquarters in regions with weaker compliance regimes. This allows them to test, scale, or monetize controversial AI systems—like surveillance analytics or data-intensive recommender algorithms—without violating stricter laws elsewhere.
-
In AI, this means systems banned in one region (e.g., EU’s high-risk classification) can still be trained offshore, then imported as models or services under a different legal label.
The result? Normative evasion—a race to the bottom where frameworks exist, but enforcement gaps make them meaningless.
Okay, what is Ethics then?
Let's understand this.
You can say, a kind of shared vocabulary that forces engineers and policymakers to stop pretending they live on different planets.
There are some basic principles of ethics, which are quite universally applicable, in the case of artificial intelligence, and even lack of jurisdiction might never be able to undo the need to address them in practice.
Transparency
Tech sees it as "can you reproduce the results?" Geopolitics sees it as "who gets to see the process?" They're not arguing—they literally mean different things by the same word.
Accountability
If an AI agent lacks technical reliability, should those who experimented it should be made as an example of "accountability" so that nobody cares to work on technical guardrails? Also, technical accountability sometimes can have economic consequences, if not legal. But markets have been hurt. What to do then?
Privacy
Tech thinks privacy is solved when data is encrypted or anonymized—a technical problem with a technical fix. Geopolitics sees privacy as "who has access and under what legal authority?"—a sovereignty problem. Engineers say "we secured the database." Diplomats ask "but which government can subpoena it?" Both think they're protecting you; neither realizes the other's solution doesn't address their threat model.
Fairness
Tech measures fairness as statistical parity across test sets—demographic groups getting equal error rates, equal opportunity, calibrated probabilities. Geopolitics asks "fair according to whom?" One jurisdiction defines discrimination by disparate impact (outcomes), another by disparate treatment (intent), and a third doesn't recognize the category at all.
Now, while it may feel that implementing these principles isn't easy, it's not impossible to think of these ideas in the most basic way as may be possible.
Let's also ask this. Do you need ethics to understand these tech & geopolitical bubbles? Absolutely.
Ethics isn’t about being moral here. It’s about translating between two dialects that don’t align — one coded in math, the other in diplomacy. When technologists and policymakers talk about “values,” they’re both describing control, just through different mediums.
Let's now understand the implementation value of AI Frameworks. Every ethical idea around AI boils down to whether it can be implemented or not.
Supervised Learning
-
Imagine a teacher giving you a math problem and the correct answer. You learn by mimicking the process.
-
How It Works: Machines are trained on labeled data (input + correct output).
-
Examples: Spam email detection, image recognition.
-
Techniques include Linear regression, decision trees, neural networks.
Unsupervised Learning
imagine being dropped into a room full of strangers and figuring out who belongs to which group based on their behaviour.
How It Works: Machines find patterns in unlabelled data.
Examples: Customer segmentation, anomaly detection.
Techniques include K-means clustering, principal component analysis (PCA).
Reinforcement Learning
Think of training a dog with treats. The dog learns which actions get rewards.
How It Works: Machines learn by trial and error through rewards and punishments.
Examples: Game-playing AIs like AlphaGo, robotics.
Techniques include Q-learning, deep reinforcement learning.
Semi-Supervised Learning
Imagine doing homework where only some answers are given. You figure out the rest based on what you know.
How It Works: Combines small labeled datasets with large unlabeled ones.
Examples: Medical image classification when labeled data is scarce.
There is a huge lack of country-specific AI Safety documentation.
Paralysis 2: Lack of Jurisdiction-Specific Documentation on AI Safety
Think of building a fire safety system for a city without knowing where fires have occurred or how they started. Without this knowledge, it’s hard to design effective safety measures.
Many countries don’t have enough local research or documentation about AI safety incidents—like cases of biased algorithms or data breaches. While governments talk about principles like transparency and privacy in global forums, they often lack concrete, country-specific data or institutions to back up these discussions with real-world evidence. This makes it harder to create effective safety measures tailored to local needs.
Neurosymbolic AI
Think of it as combining intuition (neural networks) with logic (symbolic reasoning). It’s like solving puzzles using both gut feeling and rules.
How It Works: Merges symbolic reasoning (rule-based systems) with neural networks for better interpretability and reasoning.
Examples: AI systems for legal reasoning or scientific discovery.
Here's some confession: never convert ethics terms into balloonish jargons or they won't work.
Paralysis 3: Responsible AI Is Overrated, and Trustworthy AI Is Misrepresented
Imagine a company claiming its product is "eco-friendly," but all they’ve done is slap a green label on it without making real changes. This is what happens with "Responsible AI" and "Trustworthy AI."
"Responsible AI" sounds great—it’s about accountability and fairness—but in practice, it often becomes a buzzword. Companies use these terms to look ethical while prioritizing profits over real responsibility. For example, they might create flashy ethics boards or policies that don’t actually hold anyone accountable. This dilutes the meaning of these ideals and turns them into empty gestures rather than meaningful governance.
The more garbage your questions are on AI, the more garbage will be your policy understanding on AI.
Paralysis 4: How AI Awareness Becomes Policy Distraction
Imagine everyone panicking about fixing potholes on one road while ignoring that the entire city’s bridges are crumbling. That’s what happens when public awareness drives shallow policymaking.
When people become highly aware of visible AI issues—like facial recognition—they pressure governments to act quickly. Governments often respond by creating flashy policies that address these visible problems but ignore deeper challenges like reskilling workers for an AI-driven economy or fixing outdated infrastructure. This creates a distraction from systemic issues that need more attention.
Beware: most Gen AI benchmarks are fake.
Paralysis 5: Fragmentation in the AI Innovation Cycle and Benchmarking
Imagine you’re comparing cars, but each car is tested on different tracks with different rules—one focuses on speed, another on fuel efficiency, and yet another on safety. Without a standard way to compare them, it’s hard to decide which car is actually the best. That’s the problem with AI benchmarking today.
In AI development, benchmarks are tools used to measure how well models perform specific tasks. However, not all benchmarks are created equal—they vary in quality, reliability, and what they actually measure. This practice creates confusion because users might assume all benchmarks are equally meaningful, leading to incorrect conclusions about a model’s capabilities.
Many benchmarks don’t clearly distinguish between real performance differences (signal) and random variations (noise).
A benchmark designed to test factual accuracy might not account for how users interact with the model in real-world scenarios. Without incorporating realistic user interactions or formal verification methods, these benchmarks may provide misleading assessments.
Why It Matters: Governments increasingly rely on benchmarks to regulate AI systems and assess compliance with safety standards. However, if these benchmarks are flawed or inconsistent:
Policymakers might base decisions on unreliable data.
Developers might optimise for benchmarks that don’t reflect real-world needs, slowing meaningful progress.
AI Governance priorities sometimes may not be as obvious around privacy & accountability as we know it.
Paralysis 6: Organizational Priorities Are Multifaceted and Conflicted
Imagine trying to bake a cake while three people shout different instructions: one wants chocolate frosting (investors), another wants it gluten-free (regulators), and the third wants it ready in five minutes (public trust). It’s hard to satisfy everyone.
Organizations face conflicting demands when adopting AI:
Investors want quick returns on investment (ROI) from AI projects.
Regulators require compliance with evolving laws like the EU AI Act.
The public expects ethical branding and transparency.
These competing priorities make it difficult for companies to create cohesive strategies for responsible AI adoption. Instead, they end up balancing short-term profits with long-term accountability—a juggling act that complicates governance.
Here's some truth: it never gets easy for anyone.
Paralysis 1: Regulation May or May Not Have a Trickle-Down Effect
Imagine writing a rulebook for a game, but when the players start playing, they don’t follow the rules—or worse, the rules don’t actually change how the game is played. That’s what happens when regulations fail to have the intended impact.
Governments might pass laws or policies to regulate AI, but these rules don’t always work as planned. For example, a law designed to make AI systems fairer might not actually affect how companies build or use AI because it’s too hard to enforce or doesn’t address real-world challenges. This creates a gap between policy intentions and market realities.
Still, there will be geopolitical issues around AI, and one must determine them in a reasonable way.
Start with data, and what kind of stakeholders would you need who create that resource equation.
Always remember. The geoeconomics around all AI will be more sensitive than hardcore security problems.
Why? That is a by-design reality of the digital cosmos we live in. There will also be some systemic effects that will shock you.
However, the funniest aspect of AI and geopolitics is that a typical "geoeconomic" or "economic" nexus or equation will try giving a vibe of geopolitical tensions. However, we live in a soft law world, where international rules bend more and might not be binding at all.
Another problem that may emerge is how 20th-century-based heuristics and wisdom be applied to understand the "geopolitical game", even if Systemic Effects exist such as:
-
Social inequality amplification
-
Market concentration
-
Governance or Political process interference
-
Cultural homogenisation
Instead of abstract risk categories, focus on:
Observable Impacts such as documented incidents, user complaints, system failures and performance disparities across target groups
Systemic Changes such as market structure shift, behavioural changes & cultural practice alterations in affected populations and environmental impacts
Cascading Effects such as secondary economic impacts, social relationship changes, trust in institutions and power dynamics shifts
Always ask yourself
Who is actually affected?
What changes in behaviour are we seeing?
Which impacts are measurable now?
What long-term trends are emerging?
What "geopolitical" or "geoeconomic" nexus emerging is specific to 1 kind of automation, and what is truly general enough?
Is it some old wine in a new bottle, legally, politically, economically or technologically?
But before we dive into AI Frameworks, let's take a recap to understand AI, & ML too.
Artificial Intelligence (AI) is like the term "transportation." It covers everything from bicycles to airplanes. AI refers to machines designed to mimic human intelligence—like learning, reasoning, problem-solving, and decision-making. But just as "transportation" includes many forms (cars, trains, boats), AI includes various approaches and techniques.
So, WTF is Machine Learning anyway?
Now, there are some basic concepts around artificial intelligence and geopolitics, which have stood the test of time even before the widespread use of large language models and former UK PM Boris Johnson's "chatgibbiti".
ML focuses on teaching machines to learn from data rather than being explicitly programmed. Think of it like teaching a dog tricks by showing it treats instead of manually moving its paws.
Here are some types of ML you should know.
Benchmark Capture
Imagine a university ranking that suddenly defines "success" only by test scores—but guess who makes the test? The same institutions that dominate the rankings. Benchmark Capture is when large players dictate the metrics used to judge AI reliability, safety, or fairness—creating evaluation systems they’re already optimized to win.
How It Works:
-
As Abhivardhan shows in his work on Normative Emergence, LLMs—despite being unreliable—have become the benchmark reference for all AI evaluation (citing Narayanan & Kapoor 2024; Eriksson et al. 2025). OpenAI, Anthropic, Google, and others create their own tests of factual accuracy or reasoning, but these tests aren’t scientifically grounded or cross-domain verified.
-
Smaller AI systems, or non-LLM architectures like symbolic AI or hybrid systems, are judged by standards not made for them.
This normative contagion locks the field into one family of architectures and misrepresents what “safe” or “trustworthy” AI actually means.
Perception Dysmorphia
Imagine looking in a mirror that distorts your reflection—making you see yourself as either bigger or smaller than you actually are. You make decisions based on that warped image, not reality. Perception Dysmorphia in AI governance is when policymakers, companies, and the public develop a fundamentally distorted view of what AI can do, what risks it poses, and whether governance measures are actually working—leading to regulations built on illusions rather than evidence.
How It Works:
-
Large Language Models like ChatGPT have created a false consensus about AI capabilities. Because LLMs can write fluently and mimic reasoning, people assume they're reliable, general-purpose intelligence systems. Governments then create governance frameworks based on LLM behavior—focusing on "hallucinations," "transparency," and "explainability"—and apply these norms to all AI systems, even ones that work completely differently (like computer vision, robotics, or symbolic reasoning systems).
This creates a triple distortion:
-
Overestimation: Policymakers think LLMs are more capable and trustworthy than they actually are, so they deploy them in high-stakes settings (legal advice, medical diagnosis, government services) without adequate safeguards.
-
Misapplication: Governance frameworks designed for one type of unreliable AI (LLMs) get imposed on fundamentally different AI architectures that don't share those flaws—creating regulatory mismatch.
-
Gatekeeping by Design: Compliance costs and bureaucratic requirements favor centralized AI labs with massive resources. Meanwhile, decentralized AI communities—independent developers, open-source contributors, federated learning networks—get crushed under regulations, market pressure, peer pressure, costs and maybe confusion they can't afford to manage.