In Night City, nearly anything can be upgraded. Eyes, arms, reflexes. With enough eddies, your body starts looking like an expensive collection of components with unusually aggressive warranty terms.
But there's one upgrade nobody gives you a proper user manual for: waking up with somebody else inside your head.
V didn't sign up for a roommate. Johnny Silverhand—or, more precisely, Johnny's digital engram—is not the kind of person who politely labels his food in the fridge. At the center of their conflict is the Relic, Arasaka's fictional biochip technology, and a story about digital representations of personality that becomes a struggle over identity and control.
The unsettling question isn't simply whether information can be stored. It's what happens when somebody can use that information to shape another person's choices, autonomy and future.
Cyberpunk 2077 is science fiction, not a technology forecast. But it is an excellent way to make a real governance problem impossible to ignore.
A copy of you isn't an ordinary backup
The game's world imagines a way to preserve an engram: a digital representation connected to a human mind. Johnny's engram and its presence on the Relic are central to V's story.
Nothing about that means we can currently capture consciousness, upload a person, or transfer a mind into somebody else's body. It also doesn't make an engram interchangeable with today's artificial intelligence systems.
The useful analogy lies closer to home.
Every day, real organizations process details about people: preferences, locations, behavior, messages, purchasing patterns, relationships and increasingly sophisticated inferences. Taken together, this information can create a remarkably detailed representation of an individual.
That representation isn't a mind. It can still influence what opportunities a person sees, how they're evaluated, what decisions are made about them and what others believe they know.
So we need to separate having information about a person from having legitimate authority to use it.
Data ownership, intellectual property, privacy and personal autonomy aren't one interchangeable concept. An organization can store information without owning a human being. Yet control over detailed information can create real power, especially when decisions are automated or hard to challenge.
In Night City, the story pushes that tension to a terrifying fictional extreme. In real life, it begins with something less dramatic: a product requirements meeting.
Consent is not a magic checkbox
Imagine a hypothetical service that learns from a person's messages, voice recordings and calendar to help manage daily routines. The customer chooses the assistant because it's convenient. The company later decides that the same data could be used to create a conversational avatar available to other people.
From the product team's perspective, this may look like a feature update. From a data protection perspective, it raises a question about purpose.
Under the GDPR, processing personal data about living individuals requires an appropriate lawful basis and compliance with principles such as fairness, transparency, purpose limitation, data minimization, integrity and confidentiality.
Consent can be an appropriate legal basis in some situations, but it is not required for all processing, nor does obtaining it remove every other responsibility. Where consent is relied upon, it must meet specific conditions. In other settings, a different legal basis may be relevant.
An old ticked box doesn't automatically authorize a new and unexpected use of information. An organization needs to explain what it is doing, why it is doing it, and on what basis.
The Relic doesn't give us a real legal case. It gives us a powerful question: if technology can change how someone is represented or treated, how meaningful is the control we're offering that person?
Who gets to open the file called 'you'?
Privacy discussions sometimes focus so heavily on collection notices that they forget what happens to information afterward.
Confidentiality concerns who can access it. Integrity concerns whether it can be altered without authorization or by mistake. Availability concerns whether essential information and services can be used when needed.
These aren't abstract properties when the data belongs to a person.
Return to our hypothetical assistant. If every support employee can read every private conversation, that's an access problem. If someone can manipulate the profile from which responses are generated, that's an integrity problem. If critical records become unreachable during an outage, availability becomes a problem too.
Now imagine the service sending messages to relatives while imitating its user's style, without making clear that a machine is speaking.
We haven't invented Johnny Silverhand. We have invented an obvious problem of transparency and possible deception.
Practical safeguards might include purpose-based permissions, access records, approval for particularly invasive features, change monitoring and clear distinctions between a person and a system designed to imitate aspects of them. Safeguards need to match the actual system and its impact, not the sales pitch.
The point isn't to make technology harmless. It's to make its power visible and subject to meaningful limits.
You can't outsource accountability to a chip
When something goes wrong, "the system decided" can sound like an explanation. Often, it is an attempt to avoid one.
An organization that builds or deploys tools affecting people must establish who authorizes their use, what they may do, how their behavior is evaluated, when they must be stopped, and who responds when an error causes harm.
That is accountability: decisions and safeguards that can be explained, challenged and demonstrated rather than disappearing into a vendor's black box.
This is also where AI governance becomes relevant—but not because the Relic is literally a modern AI product. It isn't. The story helps us understand why oversight, testing, documented responsibilities and human intervention matter when digital systems have significant effects on people.
The European AI Act follows a risk-based approach, with obligations for systems that fall within relevant categories and applicable provisions. Human oversight plays a specific role for certain high-risk AI systems. It is not a law about uploaded consciousness, nor does the same requirement automatically apply to every chatbot, digital avatar or piece of software that imitates a voice.
The legal analysis depends on what a system actually does, how it is used, and which rules apply at the time.
A useful operational test goes beyond putting an "Approve" button on a screen. Can the person assigned to oversee the system understand relevant warnings? Can they intervene or stop a harmful process? Are their responsibilities real—or are they expected to rubber-stamp whatever the machine produces?
Oversight without authority is just a human-shaped decoration on an automated process.
Five questions before the future ships
If Night City ran project governance workshops, they'd probably come with neon lights and an alarmingly long liability waiver. Fortunately, we can do the exercise without visiting Arasaka Tower.
Before building a system that creates detailed representations of people, ask:
What information do we actually need? Collecting more because storage is cheap is not a strategy.
Who chooses the purpose and controls access? A useful service still needs boundaries around reuse and disclosure.
How do we detect manipulation or error? A sophisticated output is not evidence that the underlying information is correct.
What can the affected person understand, question or challenge? The appropriate answers depend on the context and applicable rights.
What happens when we change or discontinue the system? A design that makes reversal impossible is a governance choice, whether or not anyone admits it.
These aren't five magic compliance checkboxes. They're conversations that help an organization make difficult choices while it still has options.
The hardest question isn't usually what the machine can do. It's what people are prepared to let it do.
Night City's most uncomfortable boundary
V and Johnny's conflict is compelling because the issue isn't merely information about a dead musician. It's a boundary: who gets to act, whose decisions matter, and who has the power to reshape somebody's deeply personal situation.
The game pushes that boundary into fiction. We face quieter versions in the real world: personal data reused beyond its purpose, unnecessarily broad access, opaque recommendations that influence consequential decisions, and responsibilities that belong to everybody on paper and nobody in practice.
One final legal distinction matters. The GDPR does not regulate digitized consciousness. Its Recital 27 also states that the Regulation does not apply to personal data of deceased persons, while Member States may provide their own rules. Philosophical questions about fictional engrams do not become legal conclusions by changing the label to "data".
But the responsibility to make real technological decisions carefully is not fictional at all.
A technology might preserve the information that describes you. Who should get to decide what happens to it?
If a platform could represent you with unsettling accuracy tomorrow, what power would you be willing to give it—and what power should never be part of the deal?
Further reading: Official Cyberpunk 2077 Ultimate Edition game booklet; EU GDPR (2016/679); EDPB basic principles; European Commission: AI Act. Original Italian article: Cyberpunk 2077: quando la tua identità diventa un dato, chi la controlla?.
Contains premise-level Cyberpunk 2077 spoilers. All real-world product examples are hypothetical; no existing technology is equivalent to the Relic.

Top comments (0)