Artificial Intelligence (AI) is already reshaping how preservation gets done, unevenly, and without the communities most affected having a say in how. Past Futures Lab is APIAHiP's initiative to build AI tools and practices for historic preservation that are accurate, community-governed, and built with Asian & Pacific Islander American and historically underrepresented communities, rather than around them.

Less than 1% of the more than 100,000 listings on the National Register of Historic Places reflect Asian & Pacific Islander American history. Combined with African American, Latino American, and Native American sites, the figure still ranges between 8-10% (National Park Service, 2013; Congressional Research Service, 2024). At the National Register's average pace of roughly a thousand new listings per year, it would take 36 to 45 years of designating exclusively BIPOC historic sites just to reflect today’s demographics, which we know continues to diversify. As of 2026, an estimated 44% of Americans identify as non-white (U.S. Census Bureau).
Preservation work, including survey, documentation, designation, project planning, fundraising and financing, advocacy, stakeholder engagement, and actual execution while complying with local, state, and federal regulations, is administratively heavy. That burden falls hardest on under-resourced communities, the same communities most likely to be documenting sites and histories that have never been formally recognized in the first place.
AI tools are entering this field already (ACHP Feb 12, 2026 Briefing PDF & YouTube). Consultants, agencies, nonprofit organizations, and community groups are using them, with no shared standards for accuracy on public history, no community consent frameworks, and no field-wide guidance on where these tools help and where they cause harm.
APIAHiP makes a deliberate choice about AI on our own communities' terms, whether that means adopting it or resisting it outright. Either way, the decision gets made intentionally. That decision takes shape in three forms: how the field learns to use these tools, how the people building them are trained, and how we test what we build before anyone else does.
01. Community Education Public workshops and resources help preservation leaders and community members understand how AI works, where it fails, and how to guide it within a preservation and heritage context. The goal is a field and a community that shapes these tools.
02. Workforce Development Community historians, preservation professionals, and AI developers meet as equals at the table. That equal footing makes real co-learning possible instead of one side teaching and the other absorbing, which keeps developers from overpromising, corporations from overselling, and everyone else grounded in what these tools can actually do.
03. Technology Innovation Past Futures Lab is APIAHiP's testing ground for ethical, community-centered digital tools. We research, prototype, and evaluate AI applications against one standard: do they make us more present with the communities we serve?

* Data Sovereignty Statement: Community histories belong to the communities that created them, not to the platforms or tools that use or store them. Past Futures Lab upholds these guidelines: (1) community ownership of cultural knowledge, (2) consent before that knowledge is used, and (3) transparency about how the knowledge is used, if it’s used at all.
** Environmental Impact Statement: We acknowledge AI's environmental cost, particularly the energy and water use at data centers. We limit our use of AI tools to text-based research, operations, and administrative support; we restrict generating images, video, and other energy-intensive AI work to specific test cases. We actively implement ways to reduce resource consumption across both our AI use and our broader operations as a fully remote organization.

Our Projects

The Past Futures Fellow is an applied research role that pairs cultural heritage practice with emerging technology. Launched in 2025, the fellowship embeds a technologist, cultural practitioner, or interdisciplinary preservationist within APIAHiP to study, prototype, and field-test AI and digital tools for underrepresented communities through:
- Project development on APIAHiP’s core technology initiatives, including ARCH, HistoryPin: East at Main Street & Decentralized Storage, APIA Everyday and other digital media platforms.
- Case studies demonstrating responsible, community-centered technology adoption and use
- Advising on ethical technology integration across APIAHiP’s programs and partnerships
- Representing APIAHiP at national convenings on heritage, planning, and technology
Past Futures Fellows
- Paul Kim [2025-2026] - Drawing on his background in landscape architecture, heritage conservation, and data analysis across the nonprofit and philanthropic sectors, Paul steers the Lab's exploration of how emerging technology can support APIA preservation practice, from strategic program design to workforce training that connects preservationists with technologists. A graduate of USC's dual-degree program in Landscape Architecture + Urbanism and Heritage Conservation, he brings a cross-sector, interdiscipinary approach to the question of what AI-informed tools owe to community-rooted preservation work.
  
  Read his thesis: Encoding Counter-Memories: Artificial Intelligence as a Tool for APIA Community Empowerment [2025]
Past Futures Lab is the workforce training and testing ground behind APIAHiP’s emerging technology initative. It is established to close the gap in the field: technologists designing AI tools for preservation often do not understand heritage and cultural ethics; preservationists often do not have the technical fluency to shape these tools sold to them.
Workshops, presentations, and both internal and external training help preservation professionals use AI and digital documentation responsibly. Cross-sector training introduces technologists to preservation ethics, data sovereignty, and cultural sensitivity. The Lab also shares open-access tools and case studies on equitable digital preservation methods by publishing our AI developments that other organizations can adopt.
ARCH is APIAHiP's first AI-assisted tool developed through the Past Futures Lab, built on models ranging from commercial platforms to fully local, open-source deployment. Its current scope is deliberately narrow: support for the National Register nomination process.
Navigating the historic designation process requires understanding overlapping federal and local regulatory frameworks, identifying the right designation pathway for a property, and producing documentation that meets formal evaluation standards. For under-resourced communities, that has usually meant hiring outside consultants.
ARCH aims to reach all three principles, while commercial AI platforms are often limited to only two:
Affordable. APIAHiP tests three deployment tiers depending on how much a project's data requires protection, from commercial frontier models to API-connected models to fully local deployment on open-source models. Which tier applies depends on data sensitivity and ownership, and vulnerability of the community and project involved.
Ethical. ARCH is community-governed. Ultimately, outputs must be reviewed by historians, preservation professionals, and community stewards and elders, while no data is shared with third-party AI systems without consent.
Useful. Within National Register nomination support, ARCH guides communities through formatting nomination forms, historic context statements, Section 106 letters, and advocacy materials, then checks accuracy and flags next steps. What used to take weeks of labor can now take hours, with professional judgment and human review central at every stage.
We are holding ARCH to this finite scope on purpose. Advocacy, public history, public programming, interpretive and preservation planning, physical site project planning, heritage tourism, and regulatory compliance are all areas where tech companies and cultural resources practitioners alike have overpromised and underdelivered on what AI can responsibly do.
Future development of the Past Futures Lab, and future iterations of ARCH, may extend into some of these areas. That expansion will happen only as far as testing, community governance, demonstrated reliability, and APIAHiP and the preservation sector’s own resource investment into emerging technologgy allow.
*ARCH does not make preservation decisions. It does not determine what is significant or what should be nominated. It does not operate without human review.
**ARCH is currently an active prototype, which you can access the public beta version below.

Open in new window/optimized for mobile:

ARCH: Agentic Resource for Cultural Heritage (public beta)

What’s Next for the Past Futures Lab?

APIAHiP is seeking partners and funders working across the sectors of heritage, history, and technology, particularly AI, to help the Lab move from research and prototyping into sustained practice. We are seeking mission-aligned, philanthropic, and trust-based relationships, and are not accepting proposals for enterprise subscriptions, affiliate marketing, or paid sponsorship and endorsement arrangements.
As funding and resources allow, APIAHiP is building toward hardware and infrastructure that we own and operate directly, such as network-attached storage (NAS) or APIAHiP-headquartered servers, so that ARCH and future AI development can run in-house rather than solely through retail-access commercial models like Anthropic's Claude, OpenAI's ChatGPT, Google's Gemini, or ChangeAgent. This is not a rejection of those platforms. Commercial models remain part of the Lab's work as a comparative testing ground, measured against locally deployed alternatives to understand tradeoffs in cost, control, and data sovereignty. We welcome infrastructural, hardware, and software partnership or collaborative opportunities on the same mission-aligned, philanthropic, and trust-based terms described above.
To date, the Lab's work has largely reached the field through conferences, webinar presentations, and select training with publications and resource documents in development. The next step is more hands-on: workshops that bring community members and technologists to the same table. We aim to train community historians and preservation professionals on how AI works and where it fails, while educating technologists and AI developers on heritage and preservation ethics. APIAHiP welcomes invitations to conferences, webinars, workshops, publications, and other direct engagement opportunities that advance this work.
The first fellowship term concludes by the end of 2026, and Paul will remain with the Lab as an advisor. While an exact timeline has not been determined for the next fellowship, APIAHiP welcomes resumes and cover letters from those interested in continuing this work, or in future Fellow, advisor, or partner opportunities more broadly. A formal position description and compensation structure for 2027+ are still in development.