AI · speech · health · responsible systems · Halifax

Building responsible AI for public benefit.

I work on natural language processing, speech, machine learning in health, responsible AI, and the practical gaps between foundations and real-world use.

Explore research View CV Contact me

Frank Rudzicz

Professor at Dalhousie · Vector Institute faculty · Killam Memorial Chair · Director of Atlantic AI Institute · Canada CIFAR Chair in AI

329 publications

29 current trainees

67 grants

130 talks

23 awards

Research

I work on language, speech, machine learning in health, and responsible AI — especially where model behaviour meets clinical, civic, and human constraints.

Human signals

Speech, language, and cognition

Speech and language as measurable, useful signals about cognition, health, access, and interaction.

Clinical systems

AI in health

Machine learning in deployment, workflow, evidence, safety, and outcomes.

Governance + evaluation

Responsible AI

Model behaviour, interpretability, fairness, privacy, standards, and public accountability.

Place + network

Atlantic capacity

Building useful AI capacity in Atlanctic Canada, for students, collaborators, public institutions, and industry.

Current

Roles and affiliations

May 2025 - present

Full professor

Faculty of Computer Science, Dalhousie University · Halifax NS

Aug 2017 - present

Faculty member

Vector Institute for Artificial Intelligence · Toronto ON

Dec 2025 - present

Professor (status)

Department of Computer Science, University of Toronto · Toronto ON

Nov 2022 - present

Affiliate Scientist

Nova Scotia Health Authority · Halifax NS

Recent

Recent work and public activity

Recent publications

2026

A Straightforward Approach to Construct ‘Lightweight’ Token Graphs from Transformers

Computational Linguistics

2026

Assessing the Quality of Mental Health Support in LLM Responses through Multi-Attribute Human Evaluation

Secure-AI4H at AAAI

2026

CausalPrism: Component-Level Evaluation of Intervention-Based Causal Reasoning in Large Language Models

EMNLP

2026

Design and User-Centred Evaluation of a Persuasive and Explainable AI App for Health Monitoring

Intelligent Systems

Awards + grants

2025

Best Paper award

GenAI4Health at IJCAI

2025

Outstanding Paper award

NAACL

2025-2029

Killam Memorial Chair

Dalhousie University

2026

Representation Immunization on Trainium: Scalable Noising & Weight-Locking

Responsible AI Amazon Research Award (ARA) · $257,710 (USD)

2026

Preventing Cognitive Atrophy in LLM Mental Health Support Through Psychometric Evaluation and Therapist-Aligned Design

OpenAI, AI and Mental Health Grant Program

2026-2031

A Unified Framework for the Safety, Interpretability, and Ethical Deployment of LLMs

NSERC Discovery · $345,000

Media

18 September 2026

CBC Radio's Information Morning

CBC Radio · radio

17 September 2026

Maritime Noon with Bob Murphy

CBC Radio · radio

14 September 2026

N.B. seeking comment on data centres while one project is already in the works

CBC News · television

2 July 2026

AI degree programs are popping up across North America. Here's where Atlantic Canada stands

CBC News · blog

9 September 2026

Gateway hubs and shared parking: AI expert has ideas on breaking Halifax’s bottlenecks

The Chronicle Herald · newspaper

Fun

Quotes, brainteasers, and Halifax notes

A small collection of quotations, puzzles, and local links.

Persons and Machines

First we shape our tools, and thereafter our tools shape us.

Marshall McLuhan

Brainteaser

Martians in a circle

A visiting logician is surrounded by a circle of Martians, each of whom either always tells the truth or always lies. She asks each Martian whether the Martian to his right is a truth teller or a liar, and from their answers, she is able to deduce the fraction of liars in the circle. What fraction is it?

Source: Winkler