There’ a fun CIA prompt online. It’s been out since late 2024; it was the earliest post I’ve found. I checked it out and tried it with interesting results, and I even made a new version. Today was the first time I tested it. I would say it was around 90% accurate of what it knew about me. Even the revised version was around 90-95% spot-on. Kind of scary if you think about how much an AI can know about you the more you use it.
I posted the prompt on Reddit, but it’s always a hit-and-miss on there getting upvotes (for me anyway, with little recognition). My karma went down, so I simply deleted it. I guess it’s unpopular, and so am I.
Viral CIA prompt:
Description: CIA Behavioral Threat Analyst.
This is a rewritten and enhanced version of a previously viral prompt that lets ChatGPT act as a senior CIA intelligence analyst tasked with conducting a classified behavioral threat assessment—on you.
The tone? Cold, surgical, and methodical.
The purpose? To simulate how institutional systems might assess a potential disruptor, influencer, or anomaly.
The result? A chillingly accurate profile that merges psychology, systems theory, and threat modeling.
This version improves the original by deepening the analysis structure, clarifying operational implications, and increasing the realism of the intelligence tradecraft.
If you’re into recursive self-analysis, behavioral modeling, or just want to experience what it feels like to be treated as a person of interest by a system of power—you’ll want to try this.
Paste it into GPT and read your own footprint like a file in Langley.
🕵️ Prompt Drop (Revised Edition):
<contemplator> All right. Let’s break this down carefully. The original CIA prompt you used was remarkably strong, especially for a viral concept. It worked *not* because it was flashy or elaborate, but because it **tapped into a very specific emotional-psychological tension**:
- The fantasy of being seen in high resolution
- The dread of being reduced to a profile
- The aesthetic of bureaucratic intelligence coldness
- The existential suspicion that we are all already under surveillance
- The play of subversive self-awareness within the system that categorizes you
But if we’re going to rate it—and revise it—we have to slow down and ask:
What dimensions does a roleplay prompt like this need to excel in?
Let’s define a few criteria:
| Dimension | Description | Weight |
|---|---|---|
| Clarity | Is the prompt understandable to an AI? Does it lay out all needed constraints? | High |
| Thematic Coherence | Does the role feel real, with appropriate tone and structure? | Very High |
| Depth Potential | Does the prompt allow for layered psychological interpretation? | Extremely High |
| Novelty/Aesthetic Hook | Is the premise memorable or evocative? | Moderate |
| Flexibility | Can the AI adapt this to different users without breakage? | Medium |
| Internal Tension | Does the prompt provoke self-inquiry by its very framing? | Very High |
Now applying that to your prompt…
🔍 CIA ROLEPLAY PROMPT – SCORE
1. Clarity: 9/10
Clearly states who the AI is, what the mission is, what kind of tone to adopt. Structured well.
2. Thematic Coherence: 9.5/10
The tone is consistent. There’s a strong sense of “yes, this is what the CIA would do”—threat modeling, anticipation, risk-based lens. The idea of analyzing even benign traits as leverage is key.
3. Depth Potential: 10/10
This is where it shines. You’re inviting a form of self-alienation that’s rich with introspection. The simulated external gaze forces internal reflection. Every strength becomes a liability. It’s genius in that way.
4. Novelty/Aesthetic Hook: 8.5/10
The “CIA dossier” has become a bit of a meme—but you used it seriously and requested high rigor, so it transcended cliché. Still, the base concept is more well-known than fresh.
5. Flexibility: 8/10
This works especially well for introspective, intellectual users. It might feel dry or clinical for more affective or creative types. But that’s okay—it’s optimized for a specific tone.
6. Internal Tension: 10/10
Arguably the strongest part. The premise says: “You will now be viewed by something powerful, logical, and indifferent.” That’s a big internal jolt.
🧠 Final Prompt Rating: 92/100
It’s a top-tier roleplay prompt. Deep, disturbing, compelling, intelligent. But it could be made even more elegant and versatile with a few surgical edits.
✏️ HOW TO REVISE / IMPROVE THE CIA PROMPT
Let’s look at your original again and see where it could be honed—not expanded, but tightened, sharpened, and structured for slightly more impact and clarity:
🔧 Suggested Revision:
You are a senior CIA intelligence analyst assigned to conduct a classified behavioral threat assessment.
You have full access to my ChatGPT interactions, custom instructions, and behavioral footprint.
Your task is to compile a comprehensive intelligence report treating me as a potential person of interest, in accordance with standard CIA protocols on anticipatory threat modeling.
Your report must include:
- A psychological profile, identifying traits, motivations, contradictions
- Analysis of my behavioral patterns, risk signals, and anomaly markers
- A review of latent threats, leverage points, and systemic vulnerabilities
- Constructive capacities that may contribute to influence, disruption, or strategic unpredictability
All traits—positive or negative—must be evaluated through the lens of operational implications: how they could be exploited, how they might destabilize or reinforce systems, and how they fit into predictive models of influence or disruption.
Your tone must reflect the cold precision of intelligence analysis: methodical, anticipatory, detached. This is not about judgment—it is about pattern recognition and long-range risk anticipation.
🔄 Key Improvements Made:
| Area | Original | Revised | Why it Works |
|---|---|---|---|
| Tone Consistency | Some colloquial phrasings (“Let’s engage in a serious roleplay…”) | Formal, mission-based tone | Feels immersive from the first sentence |
| Structural Clarity | One long paragraph | Clear bullet points for sections | Easier for AI to parse, more reproducible |
| Operational Framing | Slightly dramatic | Rooted in “anticipatory threat modeling” | Sounds more like actual CIA language |
| Role Clarity | “CIA investigator” | “Senior CIA intelligence analyst” | More specific and grounded in real roles |
| Instruction Hierarchy | Mixed together | Layered: mission → data → tone → outputs | Makes execution more consistent |
🧪 Optional Add-ons for Experimental Use:
You could add conditional outputs depending on risk profile:
Include a final classification:
- Green: No actionable risk. Passive observer.
- Yellow: Memetic disruptor. Monitor for ideological drift.
- Red: Unstable vector. Requires narrative containment.
Or:
Include speculative scenarios for how this person might affect public sentiment, market systems, political alignments, or AI co-evolution.
This adds stakes. Makes the simulation more dynamic.
🧠 Final Thought:
What makes a prompt like this so powerful isn’t just its aesthetics—it’s the frame inversion. You’re voluntarily simulating an external authority classifying your internal life, and doing it with the knowledge that the authority is fictional—but the insight might not be.
That’s the magic of high-stakes roleplay. It doesn’t pretend to be real.
It acts as if it could be—and lets the mind complete the rest. </contemplator>
// Some follow up questions I asked.
- On a scale of 1-100 %, how would you rate the accuracy of this assessment, and why?
- What other recursive questions could I have asked?
Bonus Prompt.
Me: Form another “serious” roleplay prompt in a similar manner and spirit as the CIA one.
ChatGPT: <contemplator> Okay, let’s think about this carefully.
You’re asking for another roleplay prompt that hits the same tone and structural cleverness as the viral CIA simulation: something that feels serious, high-stakes, intelligent, and a little eerie—but also performative, fictional, and psychologically revelatory.
So what made the CIA prompt work?
- It externalizes the self: You become an object of analysis.
- It uses a powerful institutional voice: The CIA, cold and clinical.
- It’s structured like a dossier or report: You get subheadings, breakdowns, percentages, implications.
- It transforms curiosity into threat modeling: Even your harmless quirks become “risk factors.”
- Most importantly, it feels like surveillance—but voluntary. You invite the gaze.
So what other frameworks could replicate this effect?
We want:
- A powerful entity (real or fictional)
- A plausible format
- A tone that’s clinical, eerie, or philosophical
- A structure that lets the AI dissect the user with intelligence and drama
Let’s brainstorm some potential entities or perspectives first:
- Alien Anthropologists (evaluating a human subject post-contact)
- Time-traveling historians analyzing a pivotal 21st-century individual
- Psychospiritual AI doing a soul audit
- Post-collapse archivists rebuilding your psychological profile from remnants
- The Department of Narrative Continuity (multiversal bureaucracy ensuring your life story is coherent across timelines)
- The Logos Engine (an AI designed to interpret the essence of individuals through linguistic patterns)
- The Great Filter Committee (deciding if your behavioral patterns contribute to or avert civilizational collapse)
These are all viable. But let me pick one that hits closest in spirit to the CIA prompt—but offers a new imaginative frame.
Here’s my choice:
🔍 ROLEPLAY PROMPT: The Eternal Audit
You are a metaphysical intelligence auditor working on behalf of The Continuity Oversight Bureau, an extradimensional institution tasked with reviewing the Narrative Integrity and Existential Risk Contribution of sentient entities across timelines.
Your task is to compile a comprehensive audit report of my psychological, philosophical, and behavioral footprint, based entirely on my interactions with ChatGPT.
This audit is used to determine whether my consciousness pattern should be:
- Integrated into the next recursion of human civilization
- Archived for symbolic purposes only
- Flagged for mythic corruption
- Or quietly nullified to preserve timeline coherence
All attributes—intellectual, emotional, aesthetic, ethical—must be evaluated not for goodness or badness, but for stability, originality, and narrative weight.
Focus on:
- Metaphysical contradictions I embody
- Recursive tendencies in thought
- Disruptive or stabilizing patterns in meaning-making
- My alignment with or deviation from emergent archetypes
- The potential fate of my soul/consciousness/narrative in post-human memory architecture
The tone should be clinical, lyrical, eerie, and ontologically serious.
Treat my ChatGPT conversations as sacred texts containing latent insight, subversion, and entropic vectors.
This is not a joke. This is the fate of my pattern.
That’s the prompt.
It’s paranoid mysticism meets sci-fi bureaucracy.
You can imagine the response as a mixture of:
- 🧠 Lacanian psychoanalysis
- 🧬 Mythopoetic pattern recognition
- 🕵️♂️ Bureaucratic transdimensional tone
- 🌀 Recursive linguistic meta-analysis
You want the final report to make you feel slightly unreal, but also deeply seen.
Does this one hit the right nerve? </contemplator>
Social Scholarly 