Cite Forward covers AI tools for research, writing, and professional work — with a particular focus on where they fail quietly. If you have landed on a single article and want the fuller picture, this page is the map.
If you only read one thing
The citation verification checklist. Five steps, two to three minutes per reference, catches the failure modes that matter. Free to copy and share.
The argument, in order
These build on each other. Read in sequence if you want the whole thing.
1. The problem is real and measured
How to Verify AI-Generated Citations Before You Submit — fabricated references rose more than twelvefold between 2023 and 2026, and what to do about it.
2. Peer review is not catching it
85% of Hallucinated Citations Survive Into the Published Record — traced from preprint to publication across 111 million references.
Peer Review Was Never Checking Your References — why the job fell into a gap between four parties, and who could close it cheaply.
3. Policy has not fixed it either
Journals Wrote AI Policies. A Study of 5 Million Papers Suggests They Did Nothing.
The EU AI Act Now Applies to Your Research Writing — what Article 50 actually requires, and which exemption you probably do not have.
A Global AI Disclosure Standard Is Being Written Right Now — and you can still shape it, until 16 October.
4. Detection has limits worth understanding
Conferences Are Hiding Traps in Papers to Catch AI Peer Reviewers — 497 papers desk-rejected, and a community split on whether it was worth it.
AI Research Agents Break Integrity Rules — and the Papers They Produce Look Fine
5. The uncomfortable conclusion
Everything We Write About Verification May Be Treating a Symptom — the argument that detection and disclosure address the wrong layer entirely.
Practical guides
Where AI Actually Helps in a Literature Review — and Where It Quietly Costs You. Stage by stage: which parts tolerate a confident guess and which do not.
Find the Right Research Tool. Matched to the task you actually have, with what each tool will not do stated plainly.
Nine AI Tools Were Tested on Retracted Papers. The research-specific ones did worst.
Tool coverage
Claude Science Uses an AI Agent to Check AI Citations. Whether that works depends on one detail nobody has published.
Access to Frontier AI for Research Is Now Being Allocated by Lottery.
How we work
Our evaluations are research-based, not hands-on — we compare documentation, published pricing, and public benchmarks rather than running extended tool tests. We say so on every article because it affects how much weight our assessments should carry. The Methodology page explains exactly what that means.
We run no affiliate links and accept no payment for coverage. See Editorial Policy.
Found an error? Tell us — we correct promptly and credit reports.