In a world that increasingly demands rapid adaptation and forward-thinking solutions, the CSA Explore Research Program aims to anticipate and address complex societal challenges.

Each research initiative within the CSA Explore Research Program comprises a collection of projects that directly contribute to advancing research on a given topic and identify how we can better prepare for emerging challenges.

CSA Group funding supports individual research projects within an initiative. Within these initiatives, recipients convene virtually several times over the project lifespan to share knowledge, discuss new ideas, and initiate new research directions.

Within the Explore Research Program, CSA Group provides strategic leadership and support, addressing common barriers associated with interdisciplinary research initiatives, including alignment, integration, funding, and knowledge mobilization.

Findings from projects are integrated and informed by strategic guidance and foresight techniques to help promote standards readiness, to identify emerging areas of importance, and to inform a broad range of parties from government, to industry, and beyond.

Current Initiatives

The CSA Explore Research Program supports research projects focused on emerging, interdisciplinary issues. We bring together researchers from any discipline that address one of these challenges, including engineering, computer science, social sciences, policy, public health, environmental sciences, and more, because complex problems require cross-disciplinary solutions. Current initiatives include:

Protecting Canada's Freshwater 

Canada’s freshwater abundance is no guarantee. A shifting climate, aging infrastructure, growing industrial demands, governance challenges, and persistent inequities are converging and accelerating to create new risks. This research initiative brings together projects that examine vulnerabilities and solutions across ecosystems, infrastructure, technology, governance, and community priorities. How do we protect and manage our freshwater sources and the people and environments that depend on them?

Securing Canada's Communications Future

Canada's communications networks are essential to daily life; even a short disruption can significantly impact healthcare, finance, transportation, and other critical systems. As climate hazards intensify and cyber threats evolve, communications systems must be resilient across physical assets, digital layers, governance, and community needs. How do we strengthen the networks that keep Canadians connected and help ensure they remain dependable in the face of emerging risks? 

About the CSA Explore Research Program and how to participate

In a world that increasingly demands rapid adaptation and forward-thinking solutions, the CSA Explore Research Program aims to anticipate and address complex societal challenges.

Each research initiative within the CSA Explore Research Program comprises a collection of projects that directly contribute to advancing research on a given topic and identify how we can better prepare for emerging challenges.

CSA Group funding supports individual research projects within an initiative. Within these initiatives, recipients convene virtually several times over the project lifespan to share knowledge, discuss new ideas, and initiate new research directions.

Within the Explore Research Program, CSA Group provides strategic leadership and support, addressing common barriers associated with interdisciplinary research initiatives, including alignment, integration, funding, and knowledge mobilization.

Findings from projects are integrated and informed by strategic guidance and foresight techniques to help promote standards readiness, to identify emerging areas of importance, and to inform a broad range of parties from government, to industry, and beyond.

Current Initiatives

The CSA Explore Research Program supports research projects focused on emerging, interdisciplinary issues. We bring together researchers from any discipline that address one of these challenges, including engineering, computer science, social sciences, policy, public health, environmental sciences, and more, because complex problems require cross-disciplinary solutions. Current initiatives include:

Protecting Canada's Freshwater 

Canada’s freshwater abundance is no guarantee. A shifting climate, aging infrastructure, growing industrial demands, governance challenges, and persistent inequities are converging and accelerating to create new risks. This research initiative brings together projects that examine vulnerabilities and solutions across ecosystems, infrastructure, technology, governance, and community priorities. How do we protect and manage our freshwater sources and the people and environments that depend on them?

Securing Canada's Communications Future

Canada's communications networks are essential to daily life; even a short disruption can significantly impact healthcare, finance, transportation, and other critical systems. As climate hazards intensify and cyber threats evolve, communications systems must be resilient across physical assets, digital layers, governance, and community needs. How do we strengthen the networks that keep Canadians connected and help ensure they remain dependable in the face of emerging risks? 

Important Dates and Deadlines

• Call for proposals: March 31, 2026

• Application deadline: June 1, 2026 11:59 pm (PDT)

• Notice of award: mid-August, 2026

• Research start date: mid-September, 2026

Value and Duration

• Funding up to $150,000 CAD for a two-year project

Eligibility Requirements

• Proposals must directly address one of the listed topics

• Principal Investigator and co-applicants must be based at a Canadian publicly funded academic institution, not-for-profit and/or public research organization

• For full details on project eligibility, please refer to the Application Guide

Apply for Funding

Please visit the CSA Group website for more details:

https://www.csagroup.org/standards/csa-research-centre/csa-explore-research-program/

We will examine emerging AI‑driven tools such as autonomous underwater vehicles, real‑time habitat monitoring systems, and machine‑learning models that reveal patterns in ocean dynamics, fisheries behavior, and ecosystem health. The talk also highlights the collaborative potential between oceanographers, technologists, and policymakers as they work to build responsible, transparent, and resilient AI systems for the marine environment.

Ultimately, this presentation argues that AI is not just a technological upgrade—it is a transformative force that can deepen our understanding of the ocean, support sustainable blue‑economy practices, and strengthen global efforts to protect the planet’s most vital and least explored realm.

AI for Narwhal and Beluga Monitoring in Northern Hudson Bay: Enhancing Conservation Through Human-in-the-Loop Artificial Intelligence

The partnership between DFO and Whale Seeker provides a compelling example of how AI can revolutionize marine mammal monitoring. Through its human-in-the-loop AI methodology, this project not only increased the accuracy and speed of analysis but also empowered decision-makers with valuable data.

So far, the Whale Seeker team has achieved:

• Detection of over 400 individuals

• Dramatic reduction in processing time

• High accuracy and adaptability under challenging environmental conditions

• Actionable insights for conservation and fisheries management

As climate change continues to impact sensitive Arctic ecosystems, AI technologies, like Whale Seeker’s Möbius tool, are building functional linkages between research, management, and action, delivering smarter, faster, and more ethical tools for a sustainable future in ocean conservation. The success of the Northern Hudson Bay project has broader implications for how AI can be deployed in ocean and wildlife monitoring more generally. Whale Seeker’s Möbius platform, and the principles behind it, are now being adapted to monitor other marine mammals, birds, and remote ecosystems around the world, including Antarctica!

This case study demonstrates that ethical, science-informed AI can enhance (not replace) human expertise, bringing both speed and quality to the pressing challenges of conservation in a changing world.

Thank you to Whale Seeker for agreeing to be featured in this Building Bridges case study.

Recently, I’ve been experimenting with using tools like ChatGPT, Claude, and Gemini, Consensus, Elicit, treating them not as "answer machines," but as sparring partners to help clear the fog. It has been a fascinating process.

I wanted to share the specific workflow and prompts that have worked for me. In the examples below, I’ll use Marine Conservation as the context, but these strategies apply to almost any field.

🚫 The Trap: "Garbage In, Garbage Out"

I used to ask simple questions like "Suggest research topics in ocean conservation," and I mostly got generic, Wikipedia-level answers. I realized that a "good" prompt needs constraints: a persona, specific context, and a clear goal.

Here are 6 Prompt Strategies I use depending on where I’m stuck. Feel free to test them!

1️⃣ The "Refiner" (When your idea is too broad)

I use this when I have a topic but no angle.

Prompt: "I am a second year PhD student in the field of [Field Name]. My broad research topic is [Topic]. Propose 5 distinct, testable research questions derived from this topic. For each question, explain why it is needed in 2026 regarding the research gap, the innovation, and the impact."

2️⃣ The "Connector" (For interdisciplinary ideas)

Use this to merge two fields (e.g., AI + Ocean Conservation).

Prompt: "I am interested in the intersection of [Topic A] and [Topic B]. Identify 5 specific, under-explored tensions or gaps between these two fields. Focus on problems that have emerged in the last 2 years."

3️⃣ The "Critic" (For testing ideas)

When I want to check for blind spots.

Prompt: "Act as a strict senior reviewer or thesis advisor in the field of [Field Name]. I am proposing the following research question: '[Insert Question]'. Critique this question. Is it too broad? Is it measurable? What is the most likely counter-argument or validity threat?"

4️⃣ The "Gap Hunter" (Finding your theoretical home)

Prompt: "Summarize the dominant theoretical frameworks used to study [Topic] in the last decade. Then, suggest an alternative or emerging theoretical lens that has not typically been applied to this topic but could yield new insights."

5️⃣ The "Architect" (The Structured Blueprint)

Use this when I need a detailed plan, not just a simple idea.

Prompt: "I’m in the early stages of planning a research project for 2026. My focus is on monitoring seal populations in Northern Canada. I want to do something impactful, not just repeat old studies. Can you help me brainstorm 5 specific, testable research questions?

For each idea, please pitch it to me by explaining:

• The Missed Opportunity: What are other researchers ignoring right now?

• The Innovation: What new tech or method would make this study stand out?

• The 'Why Now': Why is this urgent for 2026 specifically?"

6️⃣ The "Investigator" (Iterative Drilling)

I never accept the first answer. Pick the most interesting part of the AI's response and ask it to expand specifically on that.

Prompt: "In your previous response, you mentioned [Specific Concept/Variable]. Let's drill down into that. What are 3 specific ways scholars have attempted to measure this in the past, and why have those methods been insufficient? Suggest a new approach."

🔥 The "Secret Sauce": Cross-Checking (Model Stacking)

This strategy changed the game for me: Make the models fight each other. LLMs are designed to be "helpful," which means they often agree with my bad ideas. To fix this, I use a Cross-Check Loop:

1. Generate ideas in one or several AI tools (e.g., ChatGPT).

2. Paste those ideas into another AI tool (e.g. Claude) with prompts like:
   "I’m in the early stages of planning a research project for 2026. My focus is on [Topic]. One scholar suggested that I explore the question: [Suggested research topic / question from another AI tool]. Could you tell me whether this would be a strong research direction in terms of innovation, feasibility, and overall scientific value."

3. Iterate: Take the critique back to the first model and let it address these questions for you. This can be a super fun process and I usually learn from it.

⚠️ A Critical Warning

Before we commit to a research topic suggested by AI, we must understand why AI models frequently hallucinate research gaps. It comes down to some blind spots:

1. The Paywall Blindfold:
   LLM models are trained on open web data. Since a massive chunk of high-quality academic research is locked behind paywalls (Elsevier, Taylor & Francis, etc.), the AI simply cannot "see" it. It might tell you "no one has studied X," when in reality, 50 people have—but their papers are behind a login screen. LLM models often have access to metadata (titles, abstracts, keywords) because publishers make these public. It lacks in-depth understanding of the research.

2. Prediction vs. Knowledge:
   We often think AI "memorizes" facts like a database. It doesn't. It predicts the next word in a sentence based on probability. If you ask for a gap, it will construct a sentence that sounds like a plausible academic gap. It is prioritizing grammatical flow and logical sounding arguments over factual truth.

3. Simulation vs. Understanding:
   AI simulates the language of critical analysis without understanding the science. I recently listened to an interview with Yann LeCun that clarified this for me: Humans understand a simple request based on context and unspoken assumptions (like the credibility / quality of a specific journal). The machine does not. It processes information without the "lens" of a researcher.

Extended Resource:

Hall, B., & McKee, J. (2024). An early or somewhat late ChatGPT guide for librarians. Journal of Business & Finance Librarianship, 29(1), 58–69. https://doi.org/10.1080/08963568.2024.2303944

After 10+ years in research, I was curious: Can Generative AI genuinely help with high-level brainstorming, or does it just produce generic fluff?

Through trial and error, I’ve developed a workflow and specific strategies that actually work for me. I just published a full breakdown of my prompts and cross-checking strategies.

You can read the full guide here:

I’m eager to hear your experiences with these tools during the research design stage. Do you find them helpful? Do you have any go-to strategies to share?

Read the 'what we heard' report to explore what we learned.

Watch the full recording by following the link.

• Session 1: Introduction to Ocean AI and Building Bridges

• Session 2: Preparing for AI in Ocean Science

• Session 3: AI Case Study Discussions

Read the 'what we heard' report to explore what we learned.

Watch the full recording by following the link.