Digital Diplomacy Has Outgrown Social Media
On June 1, I wrote here about a shift in my own thinking: from truth to trust, and from there toward digital diplomacy. I thought I was following a question about information. I was really following a question about institutions.
For years, I had been interested in what happens when technology changes the environment in which people make sense of the world:
How do we know what is true?
Whom do we trust?
What happens when information becomes easier to produce, manipulate and distribute?
Those questions haven’t disappeared and AI has made them more urgent.
But somewhere along the way, another question began to interest me even more.
What happens to institutions when intelligence itself becomes abundant?
That question has changed how I think about digital diplomacy.
And increasingly, I think it may change what we mean by digital diplomacy altogether.
The first digital transformation happened outside the institution
For much of its recent history, digital diplomacy has been associated with communication.
Foreign ministries joined Twitter. Ambassadors developed online audiences. Embassies experimented with Instagram, Facebook, TikTok and whatever platform came next.
Diplomatic communication became faster, more direct and more public and this was a meaningful transformation.
Institutions accustomed to communicating through diplomatic notes, press briefings and carefully prepared statements suddenly found themselves operating inside a continuous global conversation.
The questions naturally followed the technology:
Should ambassadors tweet personally?
How quickly should an embassy respond?
How do governments build digital audiences?
How should diplomatic institutions measure engagement?
How do states compete for attention?
These remain useful questions, but they increasingly describe only the visible surface of a much larger transformation. Because technology is no longer sitting primarily between the foreign ministry and the public. It is moving inside the foreign ministry.
AI can translate. It can summarize diplomatic reporting. It can monitor thousands of information sources simultaneously. It can compare documents, identify patterns, draft briefings, generate scenarios and help an analyst examine in minutes what might previously have taken hours.
Tomorrow’s systems will do considerably more.
The result is easy to describe as greater efficiency.
I think something more consequential is happening.
AI is changing the scale of what one person, and eventually one institution, can do.
And once capability expands at that scale, digital diplomacy stops being mainly a communication question.
It becomes an institutional one.
Intelligence is becoming cheaper. Capability isn’t.
Imagine a diplomat working in a relatively small mission.
Until recently, the amount of information that person could monitor, translate, compare and analyze was constrained by something very human: time.
Now imagine that diplomat with increasingly capable AI systems: can follow more local sources, can translate material across languages can summarize long documents, compare policy positions, track emerging narratives, search institutional knowledge, prepare alternative scenarios, draft an initial briefing.
One diplomat can suddenly operate with analytical leverage that once required considerably more time or people. That is extraordinary! But here is the part I find more interesting.
Giving someone access to more intelligence does not automatically make the institution more capable.
The diplomat may understand a developing situation earlier than before.
Can that insight reach the right people?
Headquarters may receive more information.
Can it distinguish signal from noise?
AI may produce five plausible interpretations.
Who decides which deserves attention?
The ministry may respond faster.
Should it?
The bottleneck begins to move.
When information is scarce, obtaining intelligence is a major constraint.
When intelligence becomes abundant, other things become scarce:
attention, judgment, coordination, authority and the ability to act coherently.
That is why I suspect AI will make institutional capability more valuable, not less.
More capability can also mean more weakness
There is an assumption hiding inside many conversations about AI adoption.
If technology makes an organization more capable, giving it more technology should make it better. But institutions are rarely that simple.
Consider a ministry with weak knowledge management and give it powerful AI: the ministry may now retrieve and generate information much faster without having solved the underlying problem of what knowledge is reliable, current or institutionally authoritative.
Consider poor coordination between missions and headquarters and ddd sophisticated monitoring systems: you may create much more information moving through the same weak coordination architecture.
Consider unclear decision rights, add AI recommendations: you may accelerate decisions while making responsibility harder to locate.
This is one of the propositions I keep returning to in my work:
AI amplifies institutional capability, but it can amplify institutional weakness just as quickly.
Technology does not arrive in a vacuum. It enters organizations with histories, hierarchies, incentives, procedures, habits and existing weaknesses.
That means the effect of the same technology can be radically different from one institution to another.
And it means that the advantage may not belong to the institution that adopts the most technology. It may belong to the institution that becomes best at using technology coherently.
Imagine the first hour of a diplomatic crisis
Suppose a video appears online during an already tense bilateral dispute. It seems to show a senior government official making an inflammatory statement. Within minutes, fragments begin circulating across several platforms. Journalists start asking questions. Local political figures react. An embassy detects the video. AI-assisted monitoring tools translate reactions and identify the narratives forming around it. One system flags signs of manipulation. Another assessment is less certain.
The mission understands the local context better than anyone at headquarters. Headquarters has information the mission does not.
The communications team wants to know whether it should respond.
Political leadership needs to know whether the incident matters.
Technical specialists are still trying to establish whether the video is authentic.
Now ask a deceptively simple question:
What is the digital capability the ministry needs at this moment?
It isn’t social media. It isn’t AI. It isn’t fact-checking. It isn’t crisis communications. It is all of them working together.
The institution needs to sense what is happening, establish what it knows, recognize what it does not know, connect local context with headquarters information, determine who has authority to decide, coordinate a position and choose whether action is necessary.
Only then does the question of what to post appear.
Get Laura Iancu’s stories in your inbox
Join Medium for free to get updates from this writer.
A ministry could possess excellent technology at every point in that chain and still fail if those capabilities cannot work together.
This is why I have become increasingly interested in what we call at Diplomats.Digital capability interoperability.
Strong institutions are not defined only by the capabilities they possess.
They are also defined by whether those capabilities can function together when the institution actually needs them.
Speed changes the problem
Diplomacy creates a particularly interesting tension because diplomatic institutions were built, quite deliberately, around judgment.
Words spoken by states have consequences. Positions need coordination. Context matters. Authority matters. Sometimes ambiguity is intentional. Sometimes silence is a decision. The internet operates according to a different logic. Information moves before institutions have finished interpreting it. Synthetic content can travel faster than verification. Public expectations form before official positions do.
AI accelerates this environment again because institutions can now produce and process information at unprecedented speed.
The obvious response is to make government faster. But speed alone is a poor objective.
A ministry that makes the wrong decision in five minutes has not become more capable than one that reaches the right decision in twenty.
The real challenge is something else:
institutional coherence under speed.
When information moves faster than institutions coordinate, speed itself becomes a capability problem.
That changes how we should think about digital transformation.
The question is no longer simply: How can we make this process faster?
It becomes: What has to remain coherent as this process becomes faster?
And eventually: Who retains the authority to decide?
The authority question
This may become one of the defining institutional questions of the AI era.
AI systems will become more capable. That trajectory seems reasonably clear. They will research more. Analyze more. Recommend more. Generate more. Coordinate more. And increasingly, they will be able to execute tasks rather than merely assist with them.
The tempting question is: What can we delegate?
I think institutions need to ask another question alongside it:
What authority must we retain?
There is a profound difference between using an external system and becoming unable to function without it.
There is a difference between accepting a recommendation and no longer understanding how that recommendation was produced.
There is a difference between technological dependency and institutional dependence that can no longer be governed.
Governments have always depended on things they do not build themselves.
Dependency itself is not the problem. Dependency you cannot understand, govern, change or exit is.
This is why technological sovereignty, in practical institutional terms, cannot simply mean building everything domestically.
A foreign ministry will use external models, cloud infrastructure, software platforms, data sources and specialist vendors.
The important question is how much meaningful authority it retains while doing so.
Can the institution understand the dependency?
Can it govern how the system is used?
Can it change course?
Can it continue operating if the system becomes unavailable?
Can it leave?
Those questions sound less exciting than the latest AI capability benchmark.
For a public institution, they may matter considerably more.
And that brings the problem back to humans
There is an interesting paradox here.
The more capable machines become, the easier it is to imagine humans becoming less important. I increasingly suspect that the opposite will happen at the most consequential points.
If AI can generate ten possible analyses, someone must decide which question was worth asking.
If AI can propose five responses, someone must decide what the institution is trying to achieve.
If AI can execute increasingly complex tasks, someone must determine where execution should stop.
If systems act at machine speed, someone still carries responsibility when the consequences arrive at human speed.
As intelligence becomes abundant, agency becomes more valuable. So does judgment. So does intention.
The high-value diplomat of the AI era may therefore not be the person who knows the most about AI. It may be the person who can frame the right problem, direct increasingly abundant intelligence toward it, understand context, exercise judgment, coordinate people and systems, and take responsibility for what happens next.
Technology changes what that diplomat can do. It does not decide what that diplomat should do.
Perhaps this is the next digital divide
We often talk about technological divides between institutions that adopt new technologies quickly and those that lag behind.
AI may create a different divide. On one side will be institutions that acquire increasingly powerful tools. On the other will be institutions that develop the capabilities required to turn those tools into durable institutional advantage.
A ministry can buy technology. It cannot buy institutional coherence quite so easily.
It can deploy an AI assistant. It cannot install judgment.
It can automate a workflow. It cannot automate institutional purpose.
It can generate more intelligence. It still has to decide what matters.
This is why the work I am developing through Diplomats.Digital has increasingly moved from technology adoption toward institutional capability.
The question underneath that work is becoming quite simple:
What does this technology actually make the institution capable of doing, and what must the institution remain capable of doing for itself?
I think both halves matter.
From truth, to trust, to agency
When I wrote about moving from truth to trust in June, I thought trust might be the destination of this line of inquiry. Now I am less sure.
Trust led me toward institutions. Institutions led me toward capability. Capability led me toward authority. And authority keeps leading me back to agency. Perhaps that progression makes sense.
Technology can give us more information. AI can give us more intelligence.
Neither tells us what deserves our attention. Neither determines our purpose. Neither relieves an institution, or a person, of responsibility for choosing a direction.
That is the part of digital diplomacy I want to explore here next.
Not simply how diplomats use technology, but what happens to diplomacy when technology changes the scale of what diplomats and their institutions can know and do.
What becomes possible?
What becomes vulnerable?
What should become faster?
What should remain deliberately human?
And as more capability moves into machines, what authority should institutions never surrender?
I don’t think we have settled those questions yet. That is precisely why they are worth asking.
Technology expands possibility. Institutions determine what becomes durable capability. And humans still have to choose the direction.
