Project summary

The Waterloo AI Association (WAIA, website: https://waia.ca/) is a student-run AI safety and governance organization based in Waterloo, Canada. I founded it in January 2026, and we launched publicly on March 11th.

Although this org was founded by Laurier students, we are called WAIA because I'm targeting the Waterloo region, not just one institution. CBRE's 2025 Scoring Tech Talent report (https://www.waterlooedc.ca/blog/waterloo-rockets-into-top-10-of-cbre-tech-talent-rankings) places the Waterloo area at #7 in North America; however, there is no organized AI safety group. There are three post-secondary schools within a 20-minute radius, the University of Waterloo (41,000+ students), Wilfrid Laurier University (23,000+ students), and Conestoga College (~11,000 students). We already have a UWaterloo member and existing connections on campus, so fieldbuilding there is a continuation of our project.

Within six months:

• A co-founder and six executives, all recruited through a written application and a structured interview

• Ran 5 events

• 22 registered members (not all on our Discord)

• A website (https://waia.ca/), Discord (https://discord.gg/5JpeWwtrRb), Instagram (@waterloo.aia), and LinkedIn (https://www.linkedin.com/company/waterloo-ai-association-waia/)

• A set of short videos I built to teach the difference between AI safety and AI ethics, which we use in our own programming because that confusion is the most common thing we hit in the first conversation

We are asking for $2,000 to $5,000 to run the WAIA Fellowship, an 8-week, BlueDot-style introduction to AI safety and AI governance, and the speaker events and outreach that feed it. This is a small ask for a concrete, measurable output: two cohorts of students who have worked through a structured AI safety curriculum and know what it looks like to have a career in the field.

What are this project's goals? How will you achieve them?

Goal 1: Provide two WAIA Fellowship Cohorts of 12-20 students (Fall 2026 and Winter 2027).

The first cohort is already live. Applications close September 10th, decisions go out September 14th, and the cohort starts September 20th. This grant decides whether we run it well and whether a second one follows in winter.

The Fellowship is an 8-week program, modelled after BlueDot Impact, with one curated reading per week and one 90-minute facilitated discussion. No technical background is required. The application (https://forms.gle/gfmTSPGMDRojcGPb8) is already written.

The application is worth a look because it says something about how we are running this program. We do not score applicants on how much AI safety they already know. Prior knowledge is collected only so we can build a mixed cohort. The three things we rank on are whether someone can sit with a real uncertainty, whether they can describe a time they changed their mind, and whether they will show up for eight weeks. An intro program that admits only people who already know the material has no pipeline, which is the mistake I most wanted to avoid.

We run by conducting the intake we’ve already designed, recruiting facilitators from our existing exec team, and paying for the room, the readings, and the food that makes people actually show up.

Goal 2: Host 4-6 open events that turn curiosity into fellowship applications.

Introduction to the alignment workshop, AI safety career panel, and speaker talks. Our first target is MATS scholars, followed by Astra Fellows and Anthropic Research Fellows. They are responsive, they like to talk about their work, and it’s a win-win. Typically we run these virtually in a classroom projection, which keeps costs near zero and reaches wide. Funding includes honorariums, room logistics, and food.

Goal 3: Make WAIA the AI Safety Hub for the Waterloo Region, not just for a campus.

Our Laurier club registration was denied, and our appeal was denied too, with overlap with an existing AI-focused club given as the reason. That closed off Students' Union funding, free room bookings, and use of the university's name. It also removed the constraint that would have made this goal impossible. A registered Laurier club needs VP Clubs and Associations approval for every cross-campus collaboration and can only count fee-paying Laurier undergrads as members. As an independent organization we can put a UWaterloo student on our exec team tomorrow.

We are also federating with AI safety groups at UBC, the University of Toronto, and UW-Madison and are in conversation with the UofT group about an Ontario student organizer meetup. Funding covers transit for joint programming and for hosting organizers from other Ontario schools.

Goal 4: Actively seed a UWaterloo AI safety club and hand it to the right person when they surface.

This is the outcome I would be most excited to produce.

I am not going to run a UWaterloo club myself. I do not go there, and someone who does will always be better at it. What I can do is run the funnel that produces the person who should. The Fellowship is the mechanism: UWaterloo students come through an 8-week cohort, we watch who facilitates well and who keeps showing up, and when someone with the drive for it appears, we back them. That means handing over the operations playbook, the curriculum, the intake forms and rubric, advisor introductions, and whatever budget we have. All of that already exists in a form another person can pick up, which is why I wrote it down instead of keeping it in my head.

Whether that becomes an independent UWaterloo club or a UWaterloo chapter under WAIA is their decision. Either beats the current situation.

I cannot promise a UWaterloo club exists by April 2027, because it depends on a person I have not met yet. I can promise we will be running the program that finds them, that we already have one UWaterloo member to start from, and that when they show up, they will not have to start from zero the way I did.

Measurable outcomes at the end of the grant period:

• 8-week fellowship run with 2 cohorts of 24-40 students

• 4-6 open events run, and attendance tracked at each

• At least 3 graduates of the fellowship move on to a next step: facilitating a cohort, joining exec, applying to BlueDot or ARENA, or applying to an AI safety internship or fellowship.

• WAIA still exists in April 2027 with a leadership team that is not dependent on me personally

How will this funding be used?

The full budget assumes the maximum. At the minimum we run one cohort instead of two and cut travel.

At $5,000 (maximum):

• Fellowship programming, 2 cohorts: $1,600 (room and AV, printed reading packets, food for 16 sessions at roughly $60 per session)

• Speaker events and career panel: $1,400 (4 honorariums at $150, plus event food and logistics at roughly $200 per event)

• Outreach and recruitment: $800 (posters and QR flyers, tabling materials for the Laurier and UWaterloo club fairs, basic merch for exec visibility, mailing list tooling)

• Regional travel and federation: $900 (GO Train and transit for joint sessions with the UofT group, plus hosting an Ontario student organizer meetup)

• Contingency: $300 (absorbs venue fees and cost overruns without killing an event)

Total: $5,000

At $2,000 (minimum):

• Fellowship programming, 1 cohort: $800

• Speaker events and career panel: $600

• Outreach and recruitment: $350

• Regional travel and federation: $250

Total: $2,000

None of this money goes to me or any executive. All dollars go to programming, materials, or travel. Our operating philosophy is lean; we run world-class small-org tooling out of Google Drive and a free Notion tier, and we’d rather spend on the things that put students in a room than infrastructure.

Who is on your team? What's your track record on similar projects?

Kylie Cruz, Founder and President: Computer Science (Honours), 2nd Year, Wilfrid Laurier University. Founded WAIA in January 2026, leading it from initial concept to public launch, with an eight-person leadership team and 22 members within six months. Laurier Computing Society Outreach/Event Coordinator. ML Data Specialist (Former Intern) at AMX Research Lab. Completed BlueDot Impact’s Future of AI course. Developed a set of short videos and games that teach the difference between AI safety and AI ethics (which we use in our own programming).

King Dixon, Co-Founder & Vice President: Hired within 2 weeks of our public launch.

Executive team: Hired six additional executives in Finance, Technical, AI Safety, AI Governance, Marketing, and Outreach through a structured application and interview process with a written rubric. There is a documented handbook (https://docs.google.com/document/d/103On1geUPOxAcNShAj3Spbw5Lb93mbD6tPLd8M6QPAQ/edit?usp=sharing) for each role and a coordinator pathway underneath, so the org has a succession plan instead of a single point of failure.

Advisor: Tzu Kit Chan (Chief of Staff, Atlas Computing) has advised WAIA from early in its life and continues to be our primary outside sounding board. Our operating plan for April through August, our decision-making framework, and our approach to speaker outreach all emerged directly from those conversations.

Track record on similar projects:

• Founded and built an organization from scratch in less than three months

• Hired and onboarded eight leadership team members through a real hiring process

• We ran five events in the first semester

• Built and launched the full infrastructure stack (website, Discord w/ compliance-grade member records, Instagram, LinkedIn, member registration, and team application pipelines)

• Built an operations playbook that included exec roles, meeting cadence, decision frameworks, event runbooks, and compliance requirements, which is the thing most student groups never do and the reason most of them die in year two

• Establish federation relationships with AI safety groups at UBC, UofT, and UW-Madison

We haven't done a multi-week, structured cohort program yet. This funding is specifically for this purpose, which is why our request is modest; we are not seeking funding for the entire vision, but rather for the next step.

What are the most likely causes and outcomes if this project fails?

Most likely cause: fellowship attrition. The standard failure mode for a first cohort is 20 apply, 15 accept, and 6 finish. We’ve specifically designed the intake to accommodate this scenario. The application asks directly about the 8-week commitment, about known conflicts such as midterms and co-op terms, and about all workable time slots rather than a single preference. We want to have a smaller cohort of people that actually complete than a large one that decays. If attrition does hit us, we end up with a smaller completed cohort, not a wasted grant.

Second cause: my own capacity. I'm a full-time second-year student. Sometimes my programming slows down when my course load, a co-op term, or an internship squeezes my time. That’s why there’s a coordinator layer below each exec role as a mitigation: our AI Safety and AI Governance execs co-own the Fellowship, not just me. The fact is this org is still founder heavy, and one of my explicit goals for this grant period is to reduce that dependency.

Third cause: institutional friction. Our Wilfrid Laurier club registration was turned down, so no university room bookings, no Students’ Union account, and no institutional funding. We are now operating as an independent student organization. This situation is survivable and in some ways liberating, as an independent org can serve UWaterloo and Conestoga students without having to seek approval for every cross-campus collaboration. However, it means we pay for rooms we would otherwise obtain for free, and it’s part of our need for outside funding altogether. We will reapply for registration, and if successful, this arrangement reduces our costs and does not complicate this grant.

What failure really looks like: the downside here is minor and cheap. The money could support a couple of well-run events and a Discord community that exists but doesn’t grow. No one is harmed. The worst realistic outcome is that we produce only 12 fellowship graduates instead of 40 in a region where there are currently no fellows.

What the upside looks like: a ton of people in the Waterloo region are shipped into ML engineering roles every year, and right now they basically never run into AI safety as a serious field before they get there. A handful of students who take a formal program and take those questions into a lab or policy role is a good return on $5,000.

How much money have you raised in the last 12 months, and from where?

$0 in external funding.

WAIA has raised nothing. I have personally covered all costs to date, roughly $260 since January 2026.

We pursued two funding paths this year, and both closed:

1. Kairos Pathfinder Fellowship (Fall 2026). Applied in the spring with a written application and video component. Rejected in August 2026. The rejection was not about the strength of the group. Their stated reason was limited mentor capacity.

2. Wilfrid Laurier Students' Union club funding. Our club registration application was submitted on July 15, 2026, and denied, which closes off Students' Union budget requests, room bookings, and event funding for this year.

This Manifund application is our first external grant request and currently our only live funding path.