Detecting AI-Generated Text

Author(s)	Ms. Amala Teresa John, Ms. Sudha D
Country	India
Abstract	Nowhere near perfect, today's AI text often reads just like something a person would write. Instead of relying on one method alone, researchers now test several ways to tell machine-made words apart. Some tools learn from labeled examples, others look at number patterns hidden in sentences. Curvature in probability spaces helps spot fakes without prior training data. Hidden signals baked into outputs during generation can leave traces behind. Short stretches of text get checked individually using fine-grained models. Systems are also built to work across different topics and styles. To measure progress fairly, a new testing structure evaluates how accurate, stable, fast, and flexible each system really is. Results show most do well in controlled tests but struggle when attackers try to fool them. Real gaps remain between lab results and messy reality. Stronger defenses may come from combining steady pattern recognition with resistance to manipulation attempts.
Keywords	AI-Generated Text Detection, Large Language Models, Adversarial Robustness, Domain Generalization, Natural Language Processing.
Field	Computer > Artificial Intelligence / Simulation / Virtual Reality
Published In	Volume 7, Issue 3, May-June 2026
Published On	2026-05-02

About AIJFR Fees & Payment Current Issue Publication Archive	Submit Research Paper Track Submission Status Publication Guidelines Publication Ethics Peer Review & Plagiarism	Join as a Reviewer Editors & Reviewers Reviewer Referral Program Get Reviewer Membership Certi.	Website/Journal Policies Usage Policy Content Policies Privacy Policy

Contact Us		+91-9898-79-39-59	editor@aijfr.com

Advanced International Journal for Research