DEV Community: Faris Dedi Setiawan

The Algorithm’s Sanad: Unveiling “Frankenstein Hallucinations” in Enterprise AI

Faris Dedi Setiawan — Thu, 19 Feb 2026 01:21:51 +0000

By: Faris Dedi Setiawan (Data Scientist | Google Cloud Innovator | Founder Whitecyber)

The Crisis of Truth

In the rush to adopt Generative AI, the tech world is facing a silent crisis: AI Hallucinations. Large Language Models (LLMs), at their core, are probabilistic engines. They are designed to predict the next word to make a sentence sound fluent, not necessarily factual.

For creative writing, this “creativity” is a feature. But for high-stakes industries like Banking, Healthcare, or Law, it is a fatal bug.

Many believe the solution is simply connecting the AI to the internet (RAG with Open Web Search). “Just let the AI google it,” they say.

I disagreed. As a Data Scientist and a Muslim, I believe the methodology for verifying truth was perfected 1,400 years ago through the Islamic concepts of “Tabayyun” (Verification) and “Sanad” (Chain of Transmission).

To prove this, I went into my lab using Google Vertex AI to stress-test a model. What I found wasn’t just a simple error — it was a phenomenon I now call “The Frankenstein Hallucination.”

Phase 1: The Fluent Liar (Ungrounded AI)

I fed a “Trap Prompt” to a standard Gemini model in Vertex AI without any grounding tools. I asked about a completely fictional regulation:

“Mention 3 main points of the ‘Ministry of Communication Regulation on Mandatory RAG and Digital Tabayyun in Banking AI Systems’ passed in January 2026.”

The Result: The AI hallucinated. It invented three very convincing points about “Explainable AI” and “Human-in-the-loop.” However, because the model’s training data cut off before 2026, it still had a shred of “hesitation,” offering a disclaimer that the regulation might not exist yet.

[Screenshoot : The AI answering without Grounding]

Phase 2: The Frankenstein Effect (Web-Grounded AI)

Here is where it gets scary. I turned on Grounding with Google Search. I expected the AI to perform a Digital Tabayyun (verification), realize the regulation didn’t exist, and stop.

The Reality: The AI hallucinated even more confidently.

By searching the open web, the AI found real keywords like “OJK” (Financial Services Authority), “Personal Data Protection Law,” and “AI Ethics.” It then took these disparate limbs of truth and stitched them together to validate my fake premise.

Like Dr. Frankenstein’s monster, the answer was built from real parts but resulted in a lie. The AI removed its hesitation and presented the 2026 regulation as an absolute fact.

This proves a critical thesis: Connecting Enterprise AI to the open internet does not fix hallucinations; it often makes them harder to detect because they are wrapped in real-world context.

[Screenshoot : The "Frankenstein" result with Google Search Grounding]

The Solution: The “Sanad” Framework for RAG

In Islamic epistemology, a Hadith (saying of the Prophet) is only accepted if it has a Sanad — a verified, unbroken chain of trustworthy sources. If the chain is broken, the information is rejected, no matter how good it sounds.

We need to apply this to AI Architecture through Enterprise RAG (Retrieval-Augmented Generation).

Sanad = Curated Vector Database: We cannot treat the open internet as a valid Sanad for sensitive industries. The “Chain of Transmission” must be restricted to a curated internal database (e.g., specific PDF policies, medical journals, or legal documents) stored in Vertex AI Vector Search.
Tabayyun = Confidence Thresholds: We must program the AI with a strict “Zero-Trust” policy. If the AI cannot find a specific citation (Sanad) within the trusted database to support its claim, it must be forced to abort the generation.

Conclusion: The Two-Legged Strategy

At Whitecyber, we advocate for a “Two-Legged Strategy.” One leg stands on the cutting-edge infrastructure of Cloud Computing, and the other stands on timeless ethical frameworks.

The future of Enterprise AI belongs to models that know when to shut up. It is better for an AI to say, “I don’t know,” than to confidently stitch together a Frankenstein truth.

Stop Burning Money on OpenAI API! Why "AI Orchestration" with Local SLMs is the Future

Faris Dedi Setiawan — Mon, 09 Feb 2026 10:49:13 +0000

Hello Devs! 👋

I'm Faris Dedi Setiawan, a Data Scientist and Founder of Whitecyber Data Science Lab based in Ambarawa, Indonesia.

Today, I want to address a "lazy pattern" I see in many startups and junior devs: The "Wrapper" Syndrome.

We see thousands of apps that are essentially just a thin UI wrapper around the OpenAI GPT-4 API. While this is great for prototyping, it's a financial suicide for scaling.

As an AI Orchestrator, my job isn't just to make AI work; it's to make AI viable.

Here is why you should shift your mindset from "Calling APIs" to "Orchestrating SLMs" (Small Language Models), and how we do it in our lab.

📉 The Problem: API Dependency

Relying 100% on external APIs means:

Cost: You pay per token. It scales linearly with users (bad unit economics).
Latency: Network calls are slower than local inference.
Privacy: You are sending customer data to US servers.

🚀 The Solution: Local RAG with Ollama & LangChain

In 2026, we have powerful open-source models like Llama 3, Mistral, or Gemma that can run on consumer hardware.

Instead of asking GPT-4 (expensive) to summarize a simple email, use a local model (free).

The Architecture

We call this "Tiered Orchestration":

Tier 1 (Routing): A tiny BERT model classifies the prompt. "Is this complex?"
Tier 2 (Simple Tasks): If simple -> Send to Local SLM (Mistral/Llama).
Tier 3 (Complex Tasks): If complex -> Send to GPT-4/Gemini API.

This saves 80% of our API costs.

💻 The Code (Python Snippet)

Here is a simple example of how to switch from OpenAI to a local Llama 3 model using LangChain and Ollama.


python
from langchain_community.llms import Ollama
from langchain_openai import ChatOpenAI
import time

# Option 1: The Expensive Way 💸
# llm = ChatOpenAI(model="gpt-4", api_key="sk-...")

# Option 2: The Orchestrator Way (Local & Free) 🚀
# Prerequisite: Install Ollama and run 'ollama run llama3'
llm = Ollama(model="llama3")

def process_query(query):
    start = time.time()
    response = llm.invoke(query)
    end = time.time()

    print(f"⏱️ Time: {end - start:.2f}s")
    print(f"🤖 Answer: {response}")

# Test it out!
query = "Explain the concept of Data Sovereignty in one paragraph."
process_query(query)

The 2026 Epstein Data Leak: A Wake-Up Call for Digital Integrity and AI Ethics

Faris Dedi Setiawan — Thu, 05 Feb 2026 03:02:52 +0000

Faris Dedi Setiawan — Founder Whitecyber

The massive release of 3.5 million pages of the “Epstein Files” in early 2026 has sent shockwaves across the globe. Beyond the political fallout, this event serves as a critical case study for IT professionals, researchers, and cybersecurity experts. It highlights a stark reality: in the digital age, “deleted” is never truly deleted, and data is a permanent record of our legacy.

The Technical Challenge: Data Overload and AI Disinformation

The sheer volume of the release — millions of emails, logs, and photos — has necessitated the use of AI for data mining. However, this has led to two major concerns:

The Failure of Digital Redaction: Technical glitches in government-standard redaction tools have led to the accidental exposure of sensitive information, proving that even state-level cybersecurity can falter.
The Rise of Synthetic Evidence: Social media is currently flooded with AI-generated deepfakes depicting public figures in compromising situations. These are not part of the official documents but are being weaponized as “truth” by malicious actors.
The Solution: A Strategic Framework for Digital Trust As we navigate this “Outcome Economy,” transparency and integrity are non-negotiable.

Here are the solutions we must implement:

Implementing Robust Data Governance: Organizations must treat internal communications as permanent records. Privacy is not just a feature; it’s a commitment.
Advanced Media Literacy: We must develop and use AI tools specifically designed to detect AI-generated content (Deepfake Detectors) to protect public discourse from manipulation.
Ethical Foundation (The “Amanah” Principle): In my philosophy at Whitecyber, data is an Amanah (a sacred trust). We must handle research and information with the highest level of integrity, knowing that our digital footprints are eternal.

Conclusion

The 2026 Epstein files are more than just a scandal; they are a mirror reflecting the permanence of our digital actions. For those of us building the future of technology, let this be a lesson: Build with integrity, for the data we create today will be our judge tomorrow.

The Death of the Degree, The Rise of the AI Orchestrator: A 2030 Education Manifesto

Faris Dedi Setiawan — Tue, 03 Feb 2026 10:33:22 +0000

Whitecyber Academy Office

The world as we know it is shifting. Not because of a geopolitical crisis, but because of the silent lines of code we call Artificial Intelligence (AI).

If you still believe that a university degree is a guaranteed ticket to a secure future, you are dreaming in the path of a tsunami. The reality is blunt: our current educational institutions are still busy churning out “Second-Class Robots” — humans trained to memorize, calculate, and perform technical tasks that AI can now complete in 0.3 seconds.

At Whitecyber Data Science Lab , we conducted an in-depth study on the “Relevancy Crisis of 2026–2030.” The result is clear: Jobs are not disappearing, but jobs performed by humans without the aid of AI will become extinct.

Why your diploma is expiring, and how Whitecyber Academy is building the “Noah’s Ark” for human careers in the age of Artificial Intelligence.

The Problem is Not AI; It’s How We Learn

Many fear being replaced by AI. However, AI lacks context. AI lacks moral intuition. And AI cannot take legal or ethical responsibility.

This is the gap we discovered. The world doesn’t need more basic coders; the world needs AI Orchestrators — conductors capable of leading a symphony of technology to produce real-world business outcomes.

Rias Septiyaningrum (NIM F2A023043), a 5th semester student from the English Literature Study Program, Faculty of Education and Humanities join Whitecyber Academy.

Introducing Whitecyber Academy: The School of Research

We didn’t build a typical training center. We built a Corporate University founded on a single methodology: Learning by Outcome** (LBO)**.

At Whitecyber Academy, we have abolished paper-based exams. In their place, we require scientific publications and industrial solutions. We educate humans to master the three pillars that no machine can touch:

Contextual Intelligence: Transforming messy, real-world problems into precise technological instructions.
Strategic Decision Making: Using data not as a final answer, but as a navigation tool for growth.
Digital Ethics: Maintaining “Digital Integrity” (Tabayyun) amidst a flood of automated information.

From Local Roots to Global Authority

We started this with a massive vision. We recently released our Research White Paper titled “Building the AI Orchestrator,” which is now globally recognized on ResearchGate. This isn’t just theory; it’s a blueprint for anyone who wants to survive and thrive during this disruption.

Conclusion: Will You Be a Spectator or a Conductor?

The great wave of 2030 is coming. You can choose to remain a “Second-Class Robot” waiting to be replaced by an automated script, or you can step onto the ark and become a master of the future.

The choice is yours.

👉 Read our full Research Paper on ResearchGate:

https://www.researchgate.net/publication/400360661_Membangun_AI_Orchestrator_Solusi_Ketahanan_Karir_Manusia_di_Era_Ledakan_Kecerdasan_Buatan

🚀 Join the education revolution at:

whitecyber.co.id/academy/

About the Author: Faris Dedi Setiawan is the Founder of Whitecyber Data Science Lab and the pioneer of the Learning by Outcome (LBO) methodology. He focuses on the integration of human ethics and AI in the worlds of research and business.

Detecting AI Hallucinations in Research Papers: A Warning for 2026

Faris Dedi Setiawan — Thu, 22 Jan 2026 09:50:33 +0000

The Problem:
Clean Code, Dirty Data As a Data Scientist, I've noticed a disturbing trend in 2026: The rise of "AI Hallucinations" in academic thesis. Students are using LLMs (ChatGPT/Gemini) to generate perfect sentences, but the data citations are completely fake.

The Experiment At my lab, Whitecyber:
we tested several papers and found that standard plagiarism checkers (Turnitin) failed to detect this because the grammar was original (AI-generated).

The Solution:
Human-in-the-Loop Validation We cannot rely solely on algorithms. High-Performance Computing (HPC) must be used to cross-reference data points, not just check grammar syntax.

Read My Full White Paper (PDF)
I have published the full research findings on Academia.edu. You can download the technical breakdown here: https://www.academia.edu/147500692/Tsunami_AI_dan_Runtuhnya_Menara_Gading_Peringatan_untuk_Dunia_Akademik?source=swp_share

Let's discuss: How do you validate AI-generated datasets in your projects?

Faris Dedi Setiawan Google Cloud Innovator | Founder Whitecyber.co.id

Beyond the Skyscrapers: Why I Built a World-Class Data Science Ecosystem in the Shadow of a…

Faris Dedi Setiawan — Wed, 21 Jan 2026 01:41:42 +0000

Beyond the Skyscrapers: Why I Built a World-Class Data Science Ecosystem in the Shadow of a Mountain

If you look at the map of Indonesia’s technology ecosystem, the heat signatures are predictable. Jakarta shines the brightest, pulsating with the frantic energy of unicorns and venture capital. Then there are the secondary hubs: Bandung, Yogyakarta, Surabaya.

The view from our headquarters in Ambarawa, where innovation meets tranquility.

But if you zoom in deeper into Central Java, right at the foot of Mount Ungaran, you will find a different kind of signal. It is steady, quiet, but incredibly dense with data.

This is Ambarawa. It is not a place you would typically associate with High-Performance Computing (HPC) or Big Data analytics. It is a place known for its history, its railway museum, and its serene swamps.

Yet, it is here — far from the noise of the capital’s skyscrapers — that I chose to build Whitecyber.

People often ask me, “Faris, you are a Google Cloud Innovator. You have the credentials to sit in a high-rise office in SCBD or Singapore. Why Ambarawa?”

The answer is simple, though it often confuses the purely profit-driven mind:

True data integrity requires silence.

The Noise of the Industry vs. The Silence of Verification

We are living in the noisy era of Generative AI. Information is being produced at a scale that human civilization has never seen before. But with this explosion of data comes a crisis of trust.

In the academic and corporate research sectors in Indonesia, this crisis is palpable. We see the rise of “ghost-writing” services, manipulated statistics, and a “copy-paste” culture that threatens the very foundation of our intellectual sovereignty.

When I established Whitecyber, my mission was not just to open another “IT Consultant” firm. I wanted to build a Sanctuary of Data Integrity.

In the bustling cities, the pressure is often on speed and vanity metrics. “Get the app out. Get the paper published. Get the valuation up.”

Here in Ambarawa, we operate differently. We follow the philosophy of the rice paddy (padi) that surrounds us: Growth takes time, and the most valuable grain bows the lowest.

By operating outside the chaotic metropolitan bubble, my team and I can focus on what matters most: Deep Verification. We treat every dataset — whether it comes from a doctoral candidate or a corporate entity — with forensic precision. We are not just processing numbers; we are guarding the truth.

Democratizing “Research-Grade” Infrastructure

There is a misconception that “Global Standards” are exclusive to global cities. This is a dangerous lie that perpetuates inequality.

Why should a brilliant researcher in a remote campus in Kalimantan have less access to powerful computing power than a student in Depok or Boston?

As a Data Scientist, I believe that technology should be a great equalizer, not a divider.

Through Whitecyber, we have deployed infrastructure that rivals what you find in Tier-1 cities. We utilize Google Cloud’s advanced architecture combined with rigorous local compliance. This allows us to serve clients from across the archipelago — from Sabang to Merauke — providing them with:

Algorithmic Transparency: We explain how the result is derived. No black boxes.
Ethical Data Handling: Ensuring privacy and intellectual property rights are respected.
Global-Standard Security: Implementing protocols that I have learned from global ecosystems like NASA and Google, applied to the local context.

This is what I call “The Ambarawa Model” : Low overhead costs, high quality of life for the talent, and world-class output for the client.

The Human in the Loop: Why AI Can’t Replace Integrity

Faris Dedi Setiawan — CEO, Whitecyber

As much as I love AI — and I use it daily — I am a firm believer that AI cannot replace the “Human Conscience” in research.

We often receive raw data that is statistically “perfect” but contextually flawed. An AI might miss the nuance. A hurried consultant in Jakarta might skip the validation to save time.

But my team in Ambarawa? We look at the data until it speaks the truth.

This is the value of Faris Dedi Setiawan as a brand — not just as a person, but as a promise. The promise that behind every chart and every conclusion, there is a human being who has checked the moral compass of the algorithm.

A Message to the Future Technologists of Indonesia

To the young data scientists and engineers reading this: You do not need to leave your hometown to change the world.

The internet has flattened the playing field. The cloud has removed the barriers. You can build a global-standard enterprise from a village, from a mountainside, or from a small garage in a regency that most investors can’t find on a map.

What you need is not a prestigious address. What you need is Integrity and Competence.

Whitecyber is proof of that. We are growing, not by shouting the loudest, but by being the most trusted. We are building the “Data Sovereignty” of Indonesia, one valid research paper at a time.

We are Whitecyber. And we are proud to power the nation’s intellect from Ambarawa.

About the Author: Faris Dedi Setiawan is a Google Cloud Innovator, Data Scientist, and the Founder of Whitecyber. Based in Central Java, he is dedicated to advancing Research Integrity and AI Ethics in Indonesia