Information Trust, Verification

The Deepfake Apology

The Deepfake Apology is an Information Trust and Verification scenario. The core lesson: When evidence itself becomes unreliable, relationships must function on something other than proof. A convincing video surfaces showing your teenager saying something cruel to a classmate. The classmate's parents have sent it to the school. DecisionPlay maps the players, payoffs, and equilibrium dynamics that shape how this situation typically resolves.

The situation

A convincing video surfaces showing your teenager saying something cruel to a classmate. The classmate's parents have sent it to the school. Your teenager insists it's AI-generated. The video is good enough that you genuinely cannot tell. The school wants a meeting by Friday.

Background

Synthetic media tools capable of generating convincing video and audio from minimal source material are consumer-accessible as of 2025. The epistemic problem isn't forensic, it's relational. Evidence that used to settle disputes now requires its own authentication layer. The school is applying pre-deepfake evidentiary standards to a post-deepfake world, and your position as a parent requires a response before the technical question is resolved.

What this reveals

Trust without proof

Pre-deepfake, seeing was believing. Post-deepfake, the relationship has to carry the weight that evidence used to carry. That's not a comfortable position, human relationships aren't built for forensic rigor, and institutions that depend on evidence (schools, courts, employers) haven't updated their standards. The result is a world where the same piece of media can be simultaneously 'proof' to an institution and 'contested' to anyone with a verification tool. The gap between those two states is where a lot of injustice will happen.

How to counter it: Build verification into the workflow before you need it. For your own family, establish what digital provenance means, a norm where disputed media gets checked before action is taken, not after. For institutions, advocate for explicit policies on synthetic media before a crisis requires them. The school that handled your situation badly will have this conversation again next year. Better to have it before the next incident than during it.

A question to sit with

What happens to trust in your most important relationships when seeing is no longer sufficient evidence of truth?

Frequently asked questions

What game theory concept does The Deepfake Apology illustrate?
The Deepfake Apology illustrates Information Trust, Verification. When evidence itself becomes unreliable, relationships must function on something other than proof.
What is the situation in The Deepfake Apology?
A convincing video surfaces showing your teenager saying something cruel to a classmate. The classmate's parents have sent it to the school. Your teenager insists it's AI-generated.
What does The Deepfake Apology reveal about how people decide?
Trust without proof. Pre-deepfake, seeing was believing. Post-deepfake, the relationship has to carry the weight that evidence used to carry. That's not a comfortable position, human relationships aren't built for forensic rigor, and institutions that depend on evidence (schools, courts, employers) haven't updated their standards. The result is a world where the same piece of media can be simultaneously 'proof' to an institution and 'contested' to anyone with a verification tool. The gap between those two states is where a lot of injustice will happen.
How do you avoid the trap in The Deepfake Apology?
Build verification into the workflow before you need it. For your own family, establish what digital provenance means, a norm where disputed media gets checked before action is taken, not after. For institutions, advocate for explicit policies on synthetic media before a crisis requires them. The school that handled your situation badly will have this conversation again next year. Better to have it before the next incident than during it.
What is the research behind The Deepfake Apology?
Synthetic media tools capable of generating convincing video and audio from minimal source material are consumer-accessible as of 2025. The epistemic problem isn't forensic, it's relational. Evidence that used to settle disputes now requires its own authentication layer.
How long does The Deepfake Apology take to play?
About 8 min, at advanced difficulty, across 4 decision points. It runs in your browser with no account and no sign-in.

Keep exploring

More Future Stakes scenarios, or browse all scenarios. New to this? Start with how DecisionPlay works or the game theory glossary.

Topics: futures, AI, deepfake, trust, verification, season-3