Power does not always arrive with soldiers or sanctions.
Often, it arrives quietly — through language, norms, recommendations, and defaults.
This is soft power: the ability to shape preferences, values, and behavior without coercion. For decades, soft power flowed through movies, music, universities, media, and global institutions.
Today, it increasingly flows through artificial intelligence.
Algorithms now decide:
- What information people see
- Which voices are amplified
- Which ideas feel normal
- Which values feel reasonable
AI is becoming the most efficient soft-power instrument ever created — not because it persuades, but because it filters reality itself.
1. From Cultural Export to Cognitive Infrastructure
Traditional soft power worked downstream.
A country produced culture.
The world consumed it.
AI works upstream.
Before culture is consumed, AI systems:
- Rank content
- Translate language
- Summarize narratives
- Recommend perspectives
This means influence no longer depends only on what is produced, but on how reality is mediated.
When algorithms sit between humans and information, whoever shapes those algorithms shapes perception at scale.
Soft power becomes infrastructural.
2. Algorithms as Cultural Gatekeepers
Search engines, recommendation systems, and language models are not neutral mirrors of the world.
They are:
- Trained on selective data
- Optimized for engagement
- Designed with implicit assumptions
Every AI system answers silent questions:
- What is relevant?
- What is credible?
- What is offensive?
- What is “normal”?
These answers shape culture more efficiently than propaganda ever could — because they feel objective.
When culture is filtered algorithmically, bias does not shout.
It blends.
3. Language Models and the Power of Framing
Language is one of the deepest carriers of culture.
Large language models:
- Standardize expressions
- Normalize certain viewpoints
- Privilege dominant linguistic patterns
Languages with fewer digital resources:
- Are modeled less accurately
- Lose nuance
- Become functionally secondary
Over time, this creates linguistic gravity:
- Some ways of thinking feel more “natural”
- Others feel awkward, outdated, or invisible
This is not cultural exchange.
It is cultural re-weighting.
Related:
4. Whose Values Are Embedded in AI?
AI systems encode values through:
- Training data
- Safety rules
- Moderation policies
- Alignment choices
These values answer questions like:
- What counts as harmful?
- What speech is acceptable?
- What perspectives are balanced?
- What topics are sensitive?
These decisions are rarely democratic — yet they affect billions of users globally.
Soft power used to be negotiated through diplomacy.
Now it is deployed through default settings.
👉 Related satellite: The Myth of Neutral Algorithms
5. States, Platforms, and Narrative Advantage
Some states understand this shift better than others.
The competition is not only about building AI, but about shaping the narrative layer of the digital world.
Countries with strong AI ecosystems gain:
- Cultural reach
- Norm-setting power
- Agenda influence
Countries without them adapt to narratives shaped elsewhere.
This is why AI development is treated as a strategic priority by actors like the United States and China — while the European Union focuses primarily on rules rather than reach.
Rules constrain power.
Infrastructure projects it.
6. Recommendation Systems as Ideology Engines
Most influence today does not come from persuasion.
It comes from repetition.
Recommendation algorithms:
- Reward emotionally resonant content
- Reinforce existing beliefs
- Create feedback loops
Over time, they:
- Narrow worldviews
- Normalize extremes
- Create parallel realities
This is soft power without authorship.
No one needs to agree on a message.
The system simply amplifies what performs.
Ideology becomes emergent, not imposed.
7. AI, Soft Power, and the Global South
Soft power asymmetry mirrors AI colonial dynamics.
When AI platforms:
- Are developed elsewhere
- Reflect external values
- Moderate content by foreign norms
Local cultures face a choice:
- Adapt
- Resist
- Become invisible
Cultural sovereignty becomes difficult when:
- Language tools misinterpret context
- Moderation suppresses local discourse
- Algorithms reward foreign norms
This is not overt domination.
It is quiet cultural displacement.
👉 Related satellite: AI Colonialism
8. Education, Knowledge, and Epistemic Power
AI is rapidly becoming a learning interface:
- Explaining concepts
- Summarizing history
- Answering questions
This grants it epistemic power — influence over what people know and how they understand it.
When knowledge is mediated by AI:
- Historical framing matters
- Source selection matters
- Omission matters
Soft power here is not about persuasion.
It is about default explanation.
9. Why This Is More Powerful Than Propaganda
Propaganda requires belief.
AI does not.
It only requires use.
People do not need to trust AI fully for it to shape them.
They only need to rely on it occasionally.
And reliance compounds.
This makes AI-driven soft power:
- Persistent
- Scalable
- Resistant to criticism
You can reject propaganda.
You cannot easily reject infrastructure.
10. The Strategic Blind Spot
Many societies still debate AI as:
- A productivity tool
- A safety risk
- An ethical challenge
Few debate it as:
- A cultural force
- A narrative filter
- A soft-power engine
This blind spot favors those who move first.
By the time influence is visible, it is already normalized.
Closing Thought
Soft power once worked by attraction.
AI works by selection.
It decides what is seen, what is ignored, and what feels reasonable — long before people think they are being influenced.
When algorithms curate reality, influence no longer argues. It preselects.
That is the new face of soft power in the age of artificial intelligence.