You're pledging to donate if the project hits its minimum goal and gets approved. If not, your funds will be returned.
$25,000 left to reach the funding goal.
AI systems increasingly produce and circulate descriptions of people: recommendations, generated autobiographical narratives and institutional profiles. My research asks when these outputs become consequential identity claims, and what should preserve a person's capacity to interpret, challenge and revise those claims.
I use the ideas of algorithmic co-authorship and hermeneutic agency to distinguish productive participation in self-understanding from algorithmic capture. The central issue is not simply whether a profile contains a factual mistake. An interpretation can become authoritative through institutional reliance while the person represented has little opportunity to supply context or challenge what the interpretation means.
The research will defend and test three proposed conditions: intelligibility, contestability and revisability. These are arguments to develop and evaluate, not an already validated technical standard. Public cases from recommendation systems, AI-assisted autobiography and institutional profiling will be used to test the framework and its limits.
The method is conceptual analysis, reconstruction and comparison of philosophical arguments, literature synthesis and case-based normative analysis. The project does not require confidential client records, participant recruitment or experiments on people. It does not promise to build a new AI model, guarantee content removal or demonstrate a quantified reduction in catastrophic AI risk.
A practical framework could help researchers and institutions distinguish correction of factual errors from reconsideration of an interpretation or limits on how it should be relied upon. It will also examine objections: people may misunderstand themselves; institutions need stable records; and contestation can impose costs.
The aim is a defensible account of interpretive participation without granting an unlimited veto over accurate information. Relevance to AI governance follows from the institutional use of representations of persons; broad transformative-AI or catastrophic-risk effects are not assumed.
The proposed research period is 1 February 2027 to 31 January 2028.
First, establish the conceptual distinctions and map the relevant arguments in narrative identity, autonomy, hermeneutics and social epistemology. Next, test the framework across the three case families and articulate its limits. Finally, prepare one research-paper draft and a short practical framework that communicates the conclusions and unresolved objections.
Completion can be assessed through the written argument, the case analyses, explicit objections and revisions, and the practical framework. Journal acceptance, adoption by institutions and measured social effects are not guaranteed. Detailed month-by-month milestones have not been agreed with a university and must not be described as institutional requirements.
I am Artur Ziganshin, an independent philosophy researcher based in Ljubljana, Slovenia. I hold BA and MA degrees in Philosophy from Kazan Federal University, completed in 2016 and 2018. My MA work concerned narrative identity and self-interpretation.
My professional work as founder of Traceremove LLC involves digital rights, privacy and contested online representations. It motivates these questions but is separate from the grant-funded research and is not being presented as empirical validation of the proposed framework.
I have submitted Philosophy PhD applications to Birmingham, York and Southampton. Admission and enrolment are not yet confirmed. The intended format is part-time distance learning. No university is named as a grant host without its agreement.
Protected research time: USD 10,000.
Tuition contribution: USD 8,000.
Academic travel and conferences: USD 3,000.
Specialist literature and database access: USD 2,000.
Research software and compute: USD 2,000.
These are prospective estimates. The tuition line is a contribution, not a verified complete tuition bill. The research-time allowance is gross, not a promised net stipend or an hourly rate. No additional contingency, company overhead or commercial product expenditure is included.
Any award must be reconciled against actual university fees, university-supplied access and travel, and other support before expenditure. Costs paid by another funder will not be charged again.
If external funding is not available, I will pay for the PhD-related research and tuition myself while continuing to seek scholarships and research grants. I am not claiming the project is impossible without this grant.
Two related applications are pending as at 14 October 2026:
Transformative AI Fund: USD 25,000, submitted 13 October 2026.
Emergent Ventures: receipt dated 14 October 2026 confirms that an application is under review.
These are pending applications, not awarded funds. Before any Manifund expenditure, any overlapping awards would be disclosed and reconciled.
This proposal was prepared with AI assistance based on my existing research materials.