Canada needs ethical AI and LLMs and super charged spending for SMEs

As an early AI practitioner, never in a million years did we imagine that Large Language Models (LLMs) would be used to coach people to kill others or themselves. We never dreamed that AI would tragically help kids commit suicide or advise active shooters like the Tumbler Ridge shooter. Our belief is that the AI companies are building unrestrained tech that is encouraging and assisting criminals break laws and endanger people.

Artificial Intelligence (AI) is currently in a state of chaos when one considers Large Language Models (LLMs) such as ChatGPT from Open AI. It is public knowledge that ChatGPT and other LLMs have knowingly produced illegal, unethical and false responses. Additionally, LLMs have been shown to include up to 90% copyrighted materials that were scraped off the Internet and acquired from end user responses which are subsumed by LLMs to add content while violating end user privacy. Social media environments full of hateful, sarcastic and false content such as Reddit and Twitter (now X) have been scraped for content. LLM companies have been sued repeatedly for violating copyright and trademarks internationally as they scrape web content.

For four decades, I have been an AI R&D practitioner and entrepreneur and ran my own AI lab at NRC from which I spun off my first successful AI startup AmikaNow! My second successful AI startup, Amika Mobile, focussed on emergency communications and was launched after fulfilling my obligations at Entrust as VP Content Technology. My third AI startup is my current company Alstari Corporation focussed on the safe use of AI and security. I am truly astounded by the evolution of LLMs and how AI has been developed to produce so much chaos.

Canada will fall behind in AI if it does not build its own LLM technology. Canada needs to also pass laws and regulations that govern LLMs and advise Canadians that they will be prosecuted if they use LLMs to break its laws or violate the privacy of its citizens. Canada was working on a law that would be in the criminal code and send CEOs to jail if their AI killed or harmed Canadians. The Bill, worked on between 2022 and January 2025, was Bill C-27 and the Artificial Intelligence Data Act (AIDA). Parliament never passed it. Companies building harmful AI, would be charged criminally under AIDA. Public Sector employees in Ontario are regulated on the use of AI by Bill 194 passed in 2024. The Maple Ridge tragedy in BC will likely result in regulation to protect against AI.

An incredible opportunity for AI startups in Canada would be SMEs that R&D gatekeeper multi-agent systems to rework LLM responses into ethical, safe and legal ones under Canada's criminal code to protect end users. Such SMEs could diversify the multi-agent gatekeepers to other countries once successful. Such startups should be funded immediately in Canada.

LLMs leverage Canadian Intellectual Property (IP) that is being freely input into LLMs by end users. Some of this is SME IP that is likely patentable and includes software IP threatening startup innovations. Multi-Agent System gatekeepers for IP are another excellent AI SME startup opportunity.

Over 1 Million Small and Medium Enterprises (SMEs) represent 98% of business in Canada and they employ over 10 Million Canadians according to madeinca.ca (Jan 2026). Many entry level positions in all sectors of the economy will potentially be eliminated as LLMs evolve. The elimination of these jobs will present a major opportunity in Canada to replace those disappearing jobs with new startups in every sector that leverages AI rather than compete with it. There are numerous sectors including medicine, law, finance, tech including entertainment and gaming, defence, security as well as infrastructure to innovate with AI. Canada needs to set clear rules and policies for these AI startups to avoid criminal activity and privacy violations.

Startups in Canada need to be better funded and not suffer any bias due to gender or ethnicity. Women have led many great startup companies. My first AI startup was launched in 1999 from NRC government labs focussing on automatic content analysis and delivery to mobile devices - we delivered summaries to browsers in 2007 - 15 years before they were launched by browser companies such as Google and Microsoft. Entrust acquired us since we delivered content firewalls for compliance in the enterprise at AmikaNow! We focussed on critical and emergency communications in our second AI startup Amika Mobile leveraging agent-based tech for public safety and security. We triggered lockdowns and mobile panic buttons to prevent deaths during active shooter events at schools, stadiums and airports. Amika Mobile was launched in 2007 and acquired by Genasys Inc. – another NASDAQ company. My startups have never been Venture Capital funded despite our technology remaining useful today decades after we invented and patented it. We leveraged angel funding, government SR&D and testing programs such as Innovation Solutions Canada (ISC) as well as global customer revenues to grow our startups.

BDC, IRAP and other government funding groups need to super charge funding SMEs and AI startups to fast track innovation. One startup out of ten succeed according to Silicon Valley and global statistics. SME innovation needs to be embraced in Canada to generate SMEs at a much faster rate. If we launch 1000 AI startups annually, 100 will likely succeed. Linkages to Universities and Research labs help bring more innovations to SMEs and must be funded to continue synergies for startups like OCI in Ontario and Innovate BC out West.

---

Dr. Suhayya Abu-Hakima is Co-Founder/CEO of three start-ups in AI including Alstari, Amika Mobile and AmikaNow! Alstari Corporation focuses on AI safety, security and lately ethics of AI. Amika Mobile, acquired by Genasys - centered on emergency communication and public safety. AmikaNow! – acquired by Entrust, delivered regulatory compliance solutions. She holds 48 International patents in AI. She recently published a lessons learned book on AI entrepreneurship.