Questrade now offers tools that allow users to connect their accounts with large language models and build agentic AI trading agents, the fintech announced.
Through a new feature called Flows, users can link their Questrade accounts with AI services such as Claude and ChatGPT, describe a trading strategy in plain language, and have the platform build and execute the strategy automatically on their behalf.
Questrade is also offering a model context protocol (MCP), which connects AI models with external data sources. The protocol allows users to have Claude or ChatGPT analyze their account information and identify potential investment opportunities.
Another online trading platform, Moomoo, also has an MCP that can connect with AI services to create agentic AI trading agents.
Snap Projections launches automated recommendations
Toronto-based financial planning and analysis software Snap Projections now offers automated financial recommendations, the software company announced.
The feature can identify an optimal combination of CPP and OAS start dates, TFSA contributions, tax strategies and the order of withdrawals from registered accounts, among other planning decisions.
Other financial planning software, such as Conquest Planning, already has similar features.
Harbourfront adopts AI notetaker
Harbourfront Wealth Group is rolling out an AI notetaking platform from San Francisco-based Zocks in the coming quarter following a pilot.
Harbourfront is now Zocks’ largest Canadian customer.
During the pilot, Zocks transcribed 2,900 client meetings and handled follow-up tasks, saving advisors about 2,000 hours of work, according to Harbourfront.
Last year, Zocks added a Canadian data centre. While hosting data in Canada costs more, Zocks does not charge users extra for the option.
TD updates responsible AI principles
TD updated its responsible AI principles, applicable across the organization, the financial group announced.
The seven commitments cover compliance with the law, transparency and explainability, data use and privacy, fairness, quality and accountability, reliability and security.
TD’s AI risk management team oversees the principles, which are applied throughout the lifecycle of AI systems, from development to deployment and ongoing monitoring. AI use cases are assessed for explainability, fairness, privacy and performance before they are deployed.
“Our responsible AI principles provide a clear framework for adopting these technologies responsibly,” Ajai Bambawale, chief risk officer at TD, said in a statement.
For example, TD said its internal virtual assistants use controls including retrieval grounding (which connects an AI system to an external, verified data source), human oversight and continuous monitoring to reduce the risk of inaccurate, unsupported or inappropriate AI-generated guidance.
AI systems will continue to be monitored after deployment, while model review processes will reassess their risks and performance as they evolve, TD said.
Finance, insurance and real estate sector among leaders in workplace AI use: Statistics Canada
The use of generative AI tools at work was most prevalent in professional, scientific and technical services (65.6%), followed by finance, insurance, real estate, rental and leasing (59.2%), according to Statistics Canada.
Use was lowest in accommodation and food services (16.3%) and agriculture (17.5%).
Across all industries, the vast majority of workers (93.4%) said they were aware of generative AI tools, while about half (51.5%) said they were familiar with how the tools could be used in their current jobs.
More than two-fifths (41.6%) of workers reported using one or more AI or automation technologies as part of their work in the previous 12 months. Generative AI tools were by far the most common, used by 35.9% of all workers.
The data came from supplementary questions in Statistics Canada’s Labour Force Survey of workers aged 15 to 69 in March 2026.