You’re Not an OSINT Analyst, You’re an AI Prompt Monkey (And it Should Terrify You)
Let’s be honest, a lot of what passes for OSINT these days is simply ChatGPT rewritten and passed off as original thought. OSINT investigations are being performed by typing a question into AI, copy/pasting the answer, and calling it intelligence. This isn’t analysis, this is automation and frankly pretty risky to your credibility. If this is what your workflow looks like you are not an OSINT analyst you are a prompt monkey who is training the system to replace you.
How do I know if I am a prompt monkey?
Here is a list of a few red flags to determine if you may be an AI prompt monkey:
You don’t verify the sources AI gives you
You are unable to explain the process behind your findings
Your “intel” is based entirely on summaries
Mapping infrastructure, following pivots, and validation are not part of your process
You can’t tell whether your “intel” is a hallucination or not (or you never check)
You don’t consider human motivations behind actions
If you base your entire workflow on prompts and copy/pasting summaries of text, you’re not performing OSINT. You’re reformatting information without intelligence added. If someone asks, you should be able to explain where your data came from, how you validated the information, and why it matters.
Why This Is Dangerous Territory
Prompt monkeys are dangerous for the field of OSINT and here’s why:
Disinformation: AI has a knack for confidently summarizing and presenting bad sources as great-sounding lies.
Skill Loss/Erosion: New analysts won’t learn how to verify properly, source, write reports, or think like an adversary.
Ethical Failures: AI has no idea who it is hurting or what is legal or ethical unless trained on that information specific to the field of OSINT and law.
Replaceability: If your job is to type prompts into an AI chat, congratulations you will soon be obsolete as an analyst.
What Analysts Do That AI Can’t
This is the part of the process that no AI model (yet) can replicate:
Identifying what information is missing from the story.
Understanding context, subtext, and intent.
Questioning the lack of data as a finding.
Working within ethical boundaries that AI doesn’t understand.
Understanding human behavior, culture, and patterns over time.
Afterall, what makes great OSINT is what you question and refuse to take at face value without digging deeper.
AI Is a Tool, Not a Crutch
I’m absolutely not against the use of AI. In fact, I use it every day, I teach others how to use it, and I am even writing a book on how to apply it to your OSINT and intel investigations. AI can make simple repetitive tasks faster, summarize and outline leads, extract metadata, and even spot pivots you as a human might miss. But AI can’t always tell you what matters, it can’t take into account empathy and human emotions that drive the way people think and act, and it doesn’t care about being truthful.
"An analyst with AI is a force to be reckoned with. A prompt monkey with AI is just a paper pusher."
Time to Wake Up!
If your last report or investigation was just ChatGPT with screenshots, it’s time to get serious about your work. You don’t have to be perfect but you do have to be present. Learn how to source, understand adversarial tradecraft, hone your investigation instincts and together let’s prevent mediocrity from tarnishing the reliability of OSINT data.
Final Thoughts:
If this made you uncomfortable or upset, sit with that feeling for a bit and then reach out to the incredibly helpful OSINT community to sharpen your skills. We are all here to help you improve! AI isn’t the threat, analyst complacency is.
This is the kind of conversation I’m digging into more in my upcoming book on AI and OSINT. If you want to be the kind of analyst who survives the AI onslaught and comes out the other side as a better more well-rounded analyst- join my mailing list and stay tuned!