Editorial Independence & Conflict of Interest Policy
Effective Date: October 2026 • Last Reviewed: October 2026
1. Full Disclosure of Product Ownership (Dual Identity)
DeepResearcher.site operates both as an independent academic testing benchmark laboratory and as the creator of Research Assistant Pro (our proprietary AI literature synthesis software). Because we evaluate our own tool alongside industry peers (such as Elicit, Consensus, Scite, Perplexity, and SciSpace), we hold ourselves to a strict standard of reproducible, open empirical scrutiny.
2. Universal & Identical Benchmark Test Harness
To eliminate human reviewer bias, every tool evaluated on DeepResearcher is subjected to identical operational constraints:
- Same Prompts: All tools are queried with identical research questions, formulated across clinical medicine, computational sciences, and materials engineering.
- Identical Codebase Scoring: Leaderboard scores are never arbitrarily assigned by editors. They are computed via an open, deterministic mathematical formula in our codebase weighted across 6 academic dimensions: Citation Support (25%), Evidence Transparency (20%), Discovery (15%), Workflow (15%), Ease (10%), and Value (15%).
- Raw Output Transparency: Every benchmark run exposes raw excerpts and line-by-line DOI links, allowing readers to verify whether a citation is genuine, partially supported, or hallucinated.
3. Commercial Separation & Pay-for-Play Prohibition
- No Paid Ranking Placements: We strictly prohibit sponsored leaderboard ranks, paid score boosts, or sponsored positive reviews. No software company can pay to alter their score or placement.
- Affiliate Link Neutrality: While some outgoing links to external tools may contain standard affiliate tracking codes to cover benchmark compute and API costs, commercial relationships have zero impact on scoring algorithms or editorial verdicts.
4. Peer Review & Replication Right
AI models evolve rapidly. If a vendor releases an update or believes a specific benchmark run failed due to temporary API outage or misconfigured filters, we provide an open dispute mechanism. Anyone may submit a reproduction log via our Corrections & Dispute Center. Verified re-runs are logged and updated within 5 business days.