15/05/2026
#هام
There is a widespread misconception that publishers such as Springer and IEEE use fixed “AI percentage thresholds” to accept or reject research papers. In reality, neither Springer nor IEEE officially applies such policies.
Springer Policy on AI Use
Springer Nature journals do not publish or enforce a specific AI-detection cutoff score for manuscript decisions. While AI-detection tools may be used during editorial screening, acceptance or rejection decisions are ultimately made by editors and peer reviewers-not by automated percentage scores alone.
Springer considers AI detectors as supporting tools rather than definitive evidence of misconduct.
More importantly, Springer’s publishing policies emphasize transparency:
* AI tools such as ChatGPT must be disclosed if they were used beyond basic language editing.
* AI systems cannot be listed as authors.
* Authors remain fully responsible for the accuracy, originality, and integrity of their work.
Therefore, the primary concern is proper disclosure and research integrity, not an “AI percentage.”
Reliability of AI Detection Tools
Current research and publishing discussions acknowledge that AI-detection systems are not sufficiently reliable to justify strict thresholds. These tools can generate both false positives and false negatives.
As a result, journals generally interpret AI-detection outputs as indicators or signals-not as conclusive proof.
What Actually Matters for Springer Acceptance
Editorial evaluation focuses on:
* originality and novelty of ideas
* scientific quality and methodology
* plagiarism similarity and text overlap
* proper citation practices
* transparency regarding AI usage when applicable
IEEE and IEEE Access Policy
Similarly, IEEE and IEEE Access do not use official rules such as:
* “Below 20% AI is acceptable”
* “Above 50% AI leads to rejection”
Such percentages are not part of IEEE publishing policy.
According to IEEE publishing ethics guidelines:
1. AI tools are permitted but must be disclosed appropriately, including tools such as ChatGPT or Grammarly AI.
2. AI use alone is not treated as automatic misconduct.
3. Editorial assessment focuses on originality, plagiarism, and research integrity.
IEEE commonly uses plagiarism-detection systems such as CrossCheck to identify copied text from existing sources. These systems measure textual similarity-not “AI-generated percentage.”
Important Reality for Authors
If a manuscript submitted to IEEE Access or similar journals is flagged for potential AI-generated content, this does not automatically result in rejection.
Instead, it may lead to:
* additional editorial review
* requests for clarification or explanation
* revision or rewriting requirements
* disclosure updates
Human judgment remains the decisive factor throughout the publication process.
