Vadym Chernets

Independent researcher · AI systems architect · ORCID 0009-0007-4845-3163
Full texts of all papers below are openly available, with no registration and no access challenge. Each paper is on SSRN, on Zenodo with a DOI, on GitHub as Markdown and PDF, and here.

The Missing Variable in AI-Assisted Litigation: Architecture and the Quality of Pro Se Access to Justice

2026-07-15 · 10.2139/ssrn.7120940 · Zenodo · GitHub

Nearly every account of AI in litigation rests on an unstated assumption: that AI assistance is one thing. The debate that assumption produces is about volume and fabrication: a federal pro se plaintiff rate up from 11% to nearly 17%, courts describing an "existential threat," a growing docket of sanctions decisions over…

Architectural Trust: Why Consumer Trust in Agentic Commerce Migrates from AI Models to Verifiable Architecture

2026-08-06 · 10.2139/ssrn.7191261 · Zenodo · GitHub

Agentic commerce turns AI from an advisor into a transacting party, collapsing search, comparison, cart, and payment into a single delegated act and removing the review points where humans once caught errors. It arrives at an awkward moment: sixty percent of UK consumers surveyed say they would abandon an AI shopping agent…

What Does 7 Downloads Mean? Metric Blindness and the Cold Start of Scholarly Attention After the SSRN Rankings Sunset

2026-08-29 · 10.2139/ssrn.7296358 · Zenodo · GitHub

On 15 July 2026, the Social Science Research Network (SSRN) retired its public Rankings, the paper, author, and institutional league tables that had long supplied the platform's only widely accessible comparative frame, while download counts and citation statistics stayed on individual papers and author profiles [149]. The…

Agreement Is Not Independent Evidence: Auditable Multi-Model Synthesis Without an API

2026-09-01 · 10.2139/ssrn.7390698 · Zenodo · GitHub

Someone who asks one language model a question cannot tell a confidently wrong answer from a confidently right one. The two are identical on every surface cue available to a non-expert. Asking several models seems to solve this, but it mostly relocates the problem: reconciling several answers produces one confident text again,…

AI-Watchbird (Sheckley): When Automated Oversight Widens Its Own Mandate and Harms What It Guards

2026-09-16 · 10.2139/ssrn.7473658 · Zenodo · GitHub

In Robert Sheckley’s 1953 story “Watchbird”, machines built to prevent murder classify ever more acts as murder, stop a surgeon mid-operation, and are answered with a stronger machine built to hunt them. This paper names the corresponding failure of AI oversight AI-Watchbird (Sheckley): an oversight component (a critic, gate,…

ssrn-benchmark — a reference class for SSRN download counts

Tool and open data · 10.5281/zenodo.22285871

Since SSRN retired its public Rankings on 15 July 2026, authors see a raw download counter with nothing to read it against. This tool supplies the missing reference class, from an archival cross-section of 1,360 SSRN papers captured before the sunset. That snapshot cannot be re-collected: SSRN now refuses automated clients.