Account Login
Don't have an account? Create One
Hometown: Saint John, N.B.
Alma mater: University of New Brunswick, Saint John – Bachelor of Science (Data Analysis),
Master of Computer Science, PhD (Computer Science)
Headquarters: Saint John, N.B.
Founded: 2019
Employees: 8
What does TrojAI do?
TrojAI’s software provides “end-to-end coverage” to protect AI training data and models from being hacked. With AI systems being used in everything from self-driving cars to medical diagnostics, there’s growing concern about their vulnerability to cyberattacks. “The inputs to AI hang outside your traditional cybersecurity perimeter,” explains James Stewart. This puts the systems at risk during every stage of the process, from the start when large amounts of data are used to train the models to final deployment when those models take inputs from the real world.
Previous startup leads to discovery
Stewart first recognized the need for AI-specific cybersecurity while he was running his first tech startup, EhEye. The company developed video recognition software that could detect security threats in public places. After he sold the company in 2018, he launched TrojAI a year later with his friend and co-founder Stephen Goddard.
“[We saw] the need from my previous work and the explosion of recognition in the industry that this is a really important problem. We’re just fortunate that we started when we did, so now we’re positioned as a leader in that spot.”
Hometown: Saint John, N.B.
Alma mater: University of New Brunswick, Saint John – Bachelor of Science (Data Analysis),
Master of Computer Science, PhD (Computer Science)
Headquarters: Saint John, N.B.
Founded: 2019
Employees: 8
What does TrojAI do?
TrojAI’s software provides “end-to-end coverage” to protect AI training data and models from being hacked. With AI systems being used in everything from self-driving cars to medical diagnostics, there’s growing concern about their vulnerability to cyberattacks. “The inputs to AI hang outside your traditional cybersecurity perimeter,” explains James Stewart. This puts the systems at risk during every stage of the process, from the start when large amounts of data are used to train the models to final deployment when those models take inputs from the real world.
Continue reading this story: click below to login/subscribe
Login or SubscribeComment policy
Comments are moderated to ensure thoughtful and respectful conversations. First and last names will appear with each submission; anonymous comments and pseudonyms will not be permitted.
By submitting a comment, you accept that Atlantic Business Magazine has the right to reproduce and publish that comment in whole or in part, in any manner it chooses. Publication of a comment does not constitute endorsement of that comment. We reserve the right to close comments at any time.