AI Watch | March 2026 – New Technology

Stay on top of current issues and advances in artificial intelligence that are transforming the banking and financial services sector. Browse the latest legislative developments and emerging case law in the field. Discover focused updates relevant for you as well as broader-scale developments that are also worth a look. Our AI Watch newsletter is published periodically.

In this edition:

Canada

• The future of Canada’s AI strategy: Insights from the consultation are now published
• B.C. Court of Appeal dismisses Clearview AI’s attempt to overturn the ruling against it
• OPC joins global call for safeguards around AI‑generated imagery
• Canada and Germany sign a joint declaration on AI and launch an alliance

United States

• Financial institutions see a higher ROI when using agentic AI on accounts payable
• NIST launches the AI Agent Standards Initiative

European Union & United Kingdom 

Elsewhere in the world

• Mastercard and Visa have launched agentic payment pilots
• South Korea adopts new landmark AI regulation
• Only a minority of banks see AI-related measurable returns
• Singapore is in the lead for AI adoption in the financial sector

Canada

The future of Canada’s AI strategy: Insights from the consultation are now published

The government published a report summarizing answers submitted by the public and members of the AI Strategy Taskforce during its 30-day consultation launched last October. The consultation sought views on how Canada should update its national AI strategy. The resulting report synthesizes feedback from more than 11,000 participants.

• Respondents highlighted the importance of accelerating AI adoption across the economy while ensuring appropriate safeguards as well as governance and risk management mechanisms, such as certification standards and independent audits.
• The consultation identified a need to strengthen Canadian sovereign infrastructure, including in terms of compute, data and cloud capacity.
• Although the answers collected from respondents indicate a need for new AI regulation, the government does not state whether that is part of its upcoming AI strategy. Rumours are that Minister of AI Evan Solomon will announce a national AI strategy shortly.

B.C. Court of Appeal dismisses Clearview AI’s attempt to overturn the ruling against it

On Feb. 18, 2026, the British Columbia Court of Appeal dismissed an appeal by U.S.-based facial recognition company Clearview AI, upholding findings from the Office of the Information and Privacy Commissioner of British Columbia that the company violated B.C.’s Personal Information Protection Act (PIPA) through its collection and use of facial images without consent. This case started with investigations by privacy commissioners in British Columbia, Alberta, Québec and at the federal level, which concluded that Clearview AI’s scraping of images from the internet constituted unlawful collection of personal information. Clearview AI then filed an application for judicial review arguing that B.C.’s privacy law did not apply to its operations. The application was dismissed on the grounds that there was a real and substantial connection between the company’s activities and the province. Unsatisfied with the conclusion, Clearview AI turned to the Court of Appeal to overturn the findings against itself.

• The Court confirmed that scraping images from publicly accessible websites does not exempt organizations from consent obligations under B.C. privacy legislation, as the “publicly available” exception did not apply to Clearview’s collection of personal information. This exemption was therefore interpreted narrowly by the Court, which it justified given the quasi-constitutional status of privacy laws.
• The Court discarded the notion that Clearview AI had a right to collect personal information online that was competing with individuals’ right to personal information protection. Instead, the Court notes that it is an exercise of balancing a “need with a right.”
• Conversely, the Court of King’s Bench of Alberta concluded in a counterpart decision that although a current interpretation of “publicly available” in Alberta’s Personal Information Protection Act exemption did not cover Clearview AI’s scraping of personal information, the exemption was too restrictive of freedom of expression. As a result, the Court ordered that the exemption be broadened. However, there have not been legislative amendments since.
• As such, companies who conduct scraping activities must review their practices to align with heightened regulatory scrutiny and cannot expect to rely on the “publicly available” exemption to collect personal information online.

OPC joins global call for safeguards around AI‑generated imagery

On Feb. 23, 2026, the Office of the Privacy Commissioner of Canada joined more than 60 data protection and privacy authorities worldwide in issuing a joint statement on AI‑generated images and videos, including deepfakes. The statement highlights the significant privacy and dignity risks posed by the use of personal information in generative AI systems, particularly for children. Organizations are expected to implement robust safeguards to protection personal information, ensure transparency about the AI systems that are used, provide effective mechanisms to respond to removal requests for harmful content, and address risks specific to children.

Canada and Germany sign a joint declaration on AI and launch an alliance

On Feb. 14, 2026, Canada and Germany signed a Joint Declaration of Intent on Artificial Intelligence and announced the launch of a new Sovereign Technology Alliance, formalizing deeper bilateral cooperation on advanced technologies. Building on the Canada–Germany Digital Alliance announced in Dec. 2025, the joint declaration establishes a framework for collaboration focused on secure and sovereign AI capabilities, including compute infrastructure, research and commercialization, and talent development.

United States

Financial institutions see a higher ROI when using agentic AI on accounts payable

According to Artificial Intelligence News, unlike traditional AI projects that resulted in a ROI of 67% last year, agentic AI systems initiatives are reported to deliver an ROI of 80% by directly embedding decision‑making into operational workflows.

• Specifically, finance leaders are increasingly turning to agentic AI to deploy accounts payable (AP) automation, given that such a use case represents a great opportunity with high transaction volumes and structured data.
• For example, these AI agents can be used for invoice capture, duplicate invoice detection, fraud identification and overpayment reduction. 
• To ensure successful deployment of agentic AI in accounts payable, institutions need high‑quality, structured data and clearly defined parameters to ensure that autonomous systems can operate accurately and with appropriate oversight.

NIST launches the AI Agent Standards Initiative

The National Institute of Standards and Technology (NIST), through its Center for AI Standards and Innovation (CAISI), has launched the AI Agent Standards Initiative to support the secure and interoperable adoption of AI agents. Announced in Feb. 2026, the initiative aims to foster industry‑led technical standards and protocols that build public trust in AI agents, enable interoperability across digital systems, and promote secure deployment at scale. NIST has emphasized that while AI agents offer significant productivity benefits, their broader adoption depends on confidence in their reliability, security and ability to interact safely with other systems.

European Union & United Kingdom

Update on the EU AI Act

Concerning the Digital Omnibus aiming at simplifying the EU AI Act, the European Parliament’s co‑rapporteurs proposed minor amendments to the proposal that would restore AI literacy obligations for providers and deployers while pushing back the application of high‑risk AI obligations at least to Dec. 2, 2027.

Elsewhere in the world

Mastercard and Visa have launched agentic payment pilots

Mastercard and Visa have begun piloting “agentic payments” with major banks, testing how AI‑powered agents can execute card transactions on behalf of consumers within consent‑driven and issuer‑controlled frameworks. DBS Bank is piloting Visa’s Intelligent Commerce program in Asia Pacific, while Westpac has partnered with Mastercard’s Agent Pay framework to complete agent‑initiated card transactions in New Zealand, Australia and India. Real‑world transactions in which AI agents complete purchases using debit and credit cards have already been tested in certain sectors, and both companies are aiming to expand these agentic transactions to all transaction types. These initiatives signal a move toward automated commerce while seeking to preserve existing payment controls and trust mechanisms.

South Korea adopts new landmark AI regulation

South Korea has introduced what it describes as the world’s first comprehensive set of laws regulating artificial intelligence, with the launch of its new AI Basic Act aimed at strengthening trust and safety in AI systems. The legislation takes effect sooner than the European Union’s AI Act and is intended to support South Korea’s ambition to become one of the world’s top three AI powerhouses.

• The framework imposes specific obligations on companies deploying “high‑impact” or generative AI, including requirements for human oversight, advance user notification and clear labelling of AI‑generated content.

Only a minority of banks see AI-related measurable returns

A recent Asian Banking & Finance analysis reports that despite rapid growth in artificial intelligence investment across the banking sector, only a small minority of banks are currently realizing measurable returns. According to estimates, global banking, financial services and insurance AI spending is projected to reach US$97 billion by 2027, up sharply from US$35 billion in 2023. However, only one in ten banks reports seeing meaningful AI‑driven returns. This gap could be attributed to challenges in translating AI initiatives into operational and financial impact, highlighting that many banks continue to invest heavily in AI without clear alignment with business outcomes, governance structures or execution strategies capable of delivering sustained value.

Singapore is in the lead for AI adoption in the financial sector

According to Artificial Intelligence News, Singapore is emerging as a global leader in the financial sector for AI adoption. Nearly two‑thirds of Singapore’s financial institutions are already deploying AI in production environments, compared to much lower levels of scaled deployment globally. Specifically, common use cases amongst Singaporean financial institutions are enhancing payment technology and improving compliance and regulatory processes. Crucially, Singapore’s leading position is attributed to modernized core systems, cloud adoption, and solid governance frameworks that allow AI to be embedded safely into critical financial operations.

About BLG