The New Battlefield: Deepfakes, Artificial Intelligence and the Future of Cyberwar

At the 2026 Billington CyberSecurity Summit, conversations about deepfakes, intelligence and cyberwarfare pointed toward the same conclusion: the next national security challenge is not simply protecting information, but determining what — and whom — we can trust.

WASHINGTON, D.C. — What happens when seeing is no longer believing, cyberattacks move faster than humans can respond, and the same artificial intelligence systems creating new vulnerabilities are also necessary to defend against them?

Across several discussions at the 2026 Billington CyberSecurity Summit in Washington, D.C., speakers approached the future of national security from different directions. Deepfakes raised questions about authenticity. Intelligence officials discussed the enormous quantities of information now available to analysts. Cybersecurity experts considered how future wars could move through the digital infrastructure underlying ordinary American life.

Taken together, they described something larger.

The emerging battlefield is not purely physical or digital. It is an environment in which information, infrastructure, artificial intelligence and human judgment are becoming increasingly difficult to separate.

And at the center of it all is trust.

When Seeing Is No Longer Believing

Deepfakes are perhaps the most visible manifestation of the problem.

Artificial intelligence can now generate increasingly convincing images, audio and video. That creates obvious opportunities for deception: a fabricated statement attributed to a government official, a falsified military order, or video apparently documenting an event that never happened.

But the deeper danger is not simply that people will believe something false.

It is that people may eventually stop believing things that are true.

Once the public understands that convincing evidence can be manufactured, authentic photographs, recordings and videos can also be dismissed as AI-generated. The existence of synthetic media creates plausible deniability for genuine evidence.

That shifts the national security problem from detecting individual fakes to establishing provenance: Where did this information come from? Who created it? Has it been altered? Can its history be verified?

The contest is increasingly about controlling confidence itself.

Intelligence Has the Opposite Problem

For intelligence agencies, the challenge is not necessarily a lack of information.

It is abundance.

Satellites, sensors, commercial data, social media, open-source intelligence and other digital systems produce quantities of information that would have been unimaginable to earlier generations of intelligence officers.

The problem becomes finding what matters.

Or, as one idea from the summit captured it, finding a needle in a stack of needles.

That makes artificial intelligence indispensable. Machines can process enormous datasets, identify patterns and surface anomalies far faster than human analysts.

But the discussions also complicated the familiar question of whether AI will replace people.

The more useful model may be neither human nor machine alone.

It is the combination.

AI can search, sort and detect at extraordinary scale. Humans remain essential for context, judgment, ambiguity and understanding intent. The strategic advantage may belong to organizations that learn how to combine the two most effectively.

That same combination, however, will be available to adversaries.

Cyberwar Moves Into Everyday Life

The consequences become more tangible when the conversation turns to cyberwarfare.

Modern societies depend upon interconnected systems for electricity, water, communications, transportation, finance and government. Those connections create efficiency, but they also create vulnerabilities.

Future conflict may therefore begin long before most people recognize it as war.

A cyberattack against critical infrastructure could disrupt electricity, communications or water. A digital attack could accompany military action elsewhere in the world. An intrusion into one interconnected system could produce consequences in another that was never the original target.

Ukraine has already offered examples of cyber operations occurring alongside conventional warfare and attacks on critical infrastructure.

For Americans accustomed to thinking of war as something occurring somewhere else, cyber changes the geography of conflict.

The battlefield can reach someone's backyard.

AI Versus AI

Artificial intelligence accelerates this dynamic further.

Defenders can deploy AI systems to continuously search networks for abnormal behavior and potential intrusions. Agentic systems could eventually conduct persistent threat hunting at a scale impossible for human cybersecurity teams.

Attackers can do the same thing in reverse.

AI can search enormous systems for vulnerabilities, automate portions of attacks and adapt more rapidly than traditional cyber operations.

That means the future cybersecurity contest may increasingly involve machines probing machines — with humans operating above them, setting objectives, interpreting results and deciding when and how to act.

The important distinction is therefore not human versus AI.

It is human plus AI versus human plus AI.

The asymmetry will come from who combines them better.

The Systems We Cannot Afford to Lose

That leads to another question raised repeatedly during the summit: where should the United States invest to maintain strategic advantage?

Not every system can be defended equally.

Governments and companies will have to determine which infrastructure is truly critical, where redundancy is necessary and which systems must continue operating even after an attacker gets through.

Cybersecurity therefore cannot consist solely of building higher walls.

Organizations must assume that some attacks will succeed.

What happens next?

If the power disappears, can essential operations continue? If communications fail, is there another way to communicate? If a network becomes compromised, can an organization operate without it?

The mentality must shift from simply blocking attacks to fighting through them.

Resilience becomes a strategic capability.

The Most Valuable Target May Be Confidence

Yet physical disruption may not always be an adversary's ultimate objective.

An attacker does not necessarily have to destroy a country's infrastructure to weaken it.

Sometimes it is enough to make people unsure whether the infrastructure works.

The same principle applies to information.

A deepfake does not necessarily have to convince everyone. It may only need to introduce uncertainty. A cyberattack does not need to permanently disable a system if it can convince people that the system is unreliable.

This connects the seemingly separate problems discussed throughout the summit.

Deepfakes attack confidence in evidence.

Cyberattacks can attack confidence in infrastructure.

Information operations attack confidence in institutions.

AI dramatically increases the speed and scale at which all three can occur.

The strategic target is increasingly trust.

The Human Still Matters

There is an irony running through all of this.

The more technologically sophisticated national security becomes, the more important distinctly human capabilities may become.

AI can process extraordinary quantities of information, but humans still have to decide what matters.

Machines can detect anomalies, but humans must understand context.

Technology can establish provenance, but societies still have to decide which institutions deserve confidence.

And automated systems can defend networks at machine speed, but people must determine what those systems are permitted to do.

The future battlefield therefore may not be defined by the disappearance of humans.

It may be defined by a new division of labor between human and machine intelligence.

Cybersecurity, intelligence and information warfare are converging around the same problem: how to preserve reliable knowledge and functioning systems in an environment deliberately designed to create uncertainty.

The countries and institutions that succeed may not be those that prevent every deception or stop every intrusion. That standard may already be impossible.

They will be the ones capable of knowing what is real, recognizing what matters and continuing to operate when something inevitably goes wrong.

In the emerging security environment, trust itself is becoming critical infrastructure.

AI Transparency Disclosure: This article was developed from the author's original reporting and notes taken while attending the 2026 Billington CyberSecurity Summit in Washington, D.C. Generative AI was used to assist with organizing, synthesizing, drafting and editing the material. The reporting observations, conference notes and underlying source material originated with the author.