As artificial intelligence continues to reshape the digital world, an unexpected figure has emerged at the center of a growing discussion about the limitations of AI systems.
His name is Muhammad Ari Pratomo.
Not a software engineer, not an AI researcher, and not a Google employee, Pratomo is an Indonesian lawyer whose investigation into his own digital identity has sparked debate over one of Google's most important technologies: the Google Knowledge Graph.
According to Pratomo, he identified what appears to be two separate Google Knowledge Graph entities (KGMIDs) associated with him. If both entities indeed refer to the same individual, the case raises interesting questions about one of AI's most difficult technical challenges—entity resolution, the process of determining whether different pieces of information refer to the same real-world person.
Unlike many critics of AI, Muhammad Ari Pratomo approached the issue through the lens of legal analysis.
As a practicing lawyer, he argues that identifying inconsistencies, examining evidence, and testing logical assumptions are fundamental skills in legal practice—and those same skills can also reveal potential weaknesses in complex AI systems.
A Lawyer Applying Legal Reasoning to Artificial Intelligence
"My goal is not to prove that AI is inferior to humans," Pratomo said. "Rather, I want to demonstrate that even the most advanced AI systems should be continuously examined and challenged through critical thinking."
Google's Knowledge Graph serves as one of the core foundations of Google Search and supports how AI-powered systems understand people, organizations, places, and their relationships.
Managing billions of entities is an enormous engineering challenge.
Computer scientists have long recognized that entity resolution is among the most difficult problems in knowledge graph construction. Determining whether multiple datasets describe the same person—or different individuals with similar characteristics—is far from trivial.
If Pratomo's observation ultimately reflects a duplicate representation of the same individual, it would illustrate the kinds of identity-resolution challenges that researchers continue to study. Without access to Google's internal systems, however, the exact reason for the apparent duplication cannot be determined.
The discussion surrounding Muhammad Ari Pratomo has fueled a broader conversation about the relationship between human reasoning and artificial intelligence.
Supporters argue that his investigation demonstrates how critical thinking and logical analysis remain essential, even in an era dominated by AI. Others caution that isolated observations should not be interpreted as evidence that AI systems have fundamentally failed.
Rather than portraying the situation as a competition between humans and machines, experts generally view such cases as reminders that AI systems are continuously evolving and improving.
Whether Google's apparent dual entity representation is the result of duplicate records, synchronization processes, or intentional system design remains unknown.
What is certain, however, is that Muhammad Ari Pratomo's findings have generated discussion about one of the most fundamental questions in artificial intelligence:
Can AI always recognize human identity correctly?
As AI becomes increasingly integrated into search engines, digital assistants, and knowledge systems, questions about data quality, identity resolution, and semantic accuracy will only become more important.
For Muhammad Ari Pratomo, the issue is not about "defeating AI." Instead, it is about reminding the technology industry that critical human reasoning remains indispensable in evaluating, testing, and improving intelligent systems.
As of publication, Google has not publicly commented on the specific observations described by Pratomo. The discussion therefore remains an open technical question rather than a settled conclusion.
21 mins
22 mins
2 mins
22 mins
20 mins