Reimagining Scientific Publishing: Can AI Make Peer Review Faster, Fairer, and More Inclusive?

The structural imbalance we rarely talk about

Scientific publishing has enabled immense global progress, but its core processes have barely evolved in decades. Researchers write manuscripts, peers review them—often over months—and editorial decisions eventually follow. For those working within well-funded institutions, this lag is frustrating. But for others—particularly from lower-income regions or smaller institutions—it can be exclusionary.

What if we could preserve rigour and transparency, but reimagine the process?

A quiet experiment with AI

After completing a recent manuscript, I tested an AI model (ChatGPT-4o) to see how it would respond if asked to act as a peer reviewer. The output was remarkably detailed, well-structured, and balanced. While I wouldn’t replace human reviewers outright, I began to wonder: could this technology meaningfully assist the process?

This was the start of an idea: a new model of scientific publishing where AI is used not as a gatekeeper, but as a facilitator—providing rapid, structured feedback, and helping level the playing field for authors globally.

What this model might look like

Rather than replace human judgment, the AI would perform structured, auditable checks:

Does the manuscript adhere to accepted reporting standards?
Are the statistical methods appropriate and reproducible?
Are claims aligned with evidence provided?
Are ethical declarations, data statements, and conflicts of interest clearly disclosed?

Authors would receive this structured critique within days. Editors could use it to make informed decisions, or optionally involve human reviewers when needed. Crucially, the AI review would be published alongside the paper—transparent and accountable.

Why this matters

This isn’t about challenging established journals—they perform essential archival and curatorial functions. Instead, this idea explores an adjacent path for research that might otherwise struggle to find a platform:

Early-stage findings that are methodologically sound but not yet headline-grabbing
Work from underrepresented geographies or fields
Rigorous interdisciplinary studies that often fall outside conventional journal scopes

Most importantly, it gives authors feedback they can act on immediately, reducing the emotional and professional drain of long, uncertain review cycles.

Why I call this the “blockchain of publishing” (metaphorically)

Not because it uses literal blockchain, but because the model mirrors blockchain’s foundational values:

Transparency – Review is open and attached to the article
Decentralisation – No dependence on legacy publisher gatekeeping
Auditability – Readers can see how an article was evaluated
Participation – Future models could invite post-publication comment or community endorsement

The aim isn’t disruption for its own sake—but rather, creating trust through process, not prestige.

Cautions and caveats

Of course, many questions remain:

Can AI handle true novelty and nuance? Likely not alone—but it can be trained to support, not decide.
Will committees value research published under this model? Perhaps not yet—but they may begin to, as transparency becomes a virtue.
Could this invite misuse? It could—but thoughtful design, clear standards, and editorial oversight can help mitigate that.

As with any paradigm shift, adoption will be incremental, not immediate.

A call to collective thinking

isn’t a pitch—it’s a provocation.

Could we imagine a world where scientific review is faster, fairer, and more inclusive—without compromising rigour?
Where technology supports judgment, but doesn’t replace it?
Where publication is based on transparency, not prestige signals alone?

I’d love to hear your thoughts—especially from those who’ve felt excluded, frustrated, or inspired by the current system.

Let’s explore what’s possible—together.

The views in this article are my own and intended to foster constructive dialogue within the academic and scientific communities.