Principle

DeshiApp may use artificial intelligence and automated tools to assist research, translation, data analysis, and copy refinement, but human editorial agency remains absolute.

Every published article remains attributable to an authenticated human author or responsible editor. Algorithmic systems possess no ethical responsibility or legal accountability.

Scope

This policy applies to generative text systems, automated code assistants, translation models, machine transcription services, synthetic media generators, and algorithmic indexing tools across all DeshiApp channels.

Permitted Assistance

AI may assist brainstorming, preliminary topic mapping, transcription of recorded interviews, grammatical proofreading, code debugging, and structuring raw datasets under human verification.

Permission to use a tool does not remove the need for human verification. Authors remain personally responsible for every sentence, number, citation, and attribution published.

Prohibited Uses

DeshiApp does not use AI to fabricate a source, interview, document, benchmark, or quotation. We strictly forbid generating unvetted bulk articles or deploying synthetic author personas.

Fluency, confidence, and grammatical quality are not evidence of factual accuracy. Machine-generated text must never be accepted without independent empirical corroboration.

AI Output Is Not Evidence

Every AI-generated factual claim is checked against reliable non-AI primary evidence, recognizing that language models frequently hallucinate facts, invent citations, and reflect systemic training bias.

AI-generated source lists are treated as leads rather than verified bibliographies. Citations must be confirmed directly against original published papers and records.

Material Disclosure

AI use is disclosed when it would reasonably affect a reader’s assessment of the work, such as substantive data clustering, automated translation, or AI-generated conceptual illustrations.

Routine spellcheck, punctuation suggestions, formatting, and standard search engine queries do not require public disclosure.

Disclosure Language

A disclosure identifies what the tool was used for in plain, descriptive language—such as ‘Audio transcribed using automated speech recognition and verified by editors’.

Naming a tool alone is not an adequate disclosure. Readers must be informed of the functional scope and degree of human oversight applied.

Authorship and Identity

AI systems are not credited as authors. A byline is reserved exclusively for a real, accountable human writer with verifiable professional credentials.

AI-Assisted Research and Drafting

AI may help identify questions, terminology, or possible perspectives, but must not replace substantive human reporting, investigative interviews, or independent critical analysis.

AI-assisted drafting remains under human control. Authors must ensure published prose reflects original human intellect rather than mechanical algorithmic regurgitation.

Translation and Transcription

Automated translation may assist preliminary understanding, but critical quotes, legal terms, and technical nuances require human verification by a qualified linguist.

Automated transcription is checked against original audio recordings before quotes are published, ensuring technical terms and cultural expressions are captured accurately.

Data Analysis and Coding

AI may assist data extraction, classification, pattern discovery, and script writing, provided underlying algorithms and diagnostic scripts are audited and documented.

AI-generated code may contain security weaknesses, unoptimized queries, or flawed logic. All code snippets published or used in research must be tested in secure environments.

Synthetic Images

Synthetic editorial images are labeled when a reasonable reader could otherwise mistake them for authentic photographs, screenshots, or documentary evidence.

Illustrative synthetic images may be used for conceptual topics, but documentary reporting and software reviews require authentic, unmanipulated screenshots.

Synthetic Audio, Voice and Video

Synthetic or cloned voices create high risks of deception and impersonation, and are not deployed in DeshiApp productions except when analyzing synthetic voice technology itself.

A reconstruction is based on identified evidence sources and is labeled prominently so audiences clearly recognize the synthetic nature of the presentation.

Detection and Reporting on AI

Synthetic-media detection tools may assist evaluation, but are not treated as conclusive proof given known false-positive and false-negative rates.

Reporting on AI systems distinguishes demonstration benchmarks from real-world performance, examining compute requirements, commercial incentives, and algorithmic safety.

Confidentiality and Personal Data

Confidential, personal, legally privileged, embargoed, or proprietary material must never be uploaded to third-party public AI tools where security controls are uncertain.

AI use involving personal data is limited to what is lawful and necessary under applicable privacy legislation, including the DPDP Act and relevant data frameworks.

Copyright and Commercial Content

AI-assisted content may create copyright, licensing, and fair dealing complexities. Writers ensure all referenced materials respect intellectual property rights.

Commercial content using AI is subject to the same disclosure standards as editorial work, ensuring marketing materials never mislead consumers regarding AI capabilities.

Automated Moderation and Recommendations

If DeshiApp introduces automated filtering or recommendation algorithms, human moderation oversight remains active to prevent algorithmic bias or arbitrary censorship.

Corrections and Contributor Declaration

AI-related errors are corrected under the Corrections Policy, with public notices acknowledging the role of automation where an algorithmic tool contributed to the failure.

Contributors declare material AI use before publication, confirming that all factual assertions were independently verified against primary human sources.

Contact

Questions concerning AI use or reports of potential synthetic media misattributions may be sent to editorial@deshiapp.in. Formal complaints should be directed to complaints@deshiapp.in.