CapabilitiesMaturityApproachDiagnosticAI & Tech ReviewContact
Weekly Review

AI & Tech Review

Emerging developments across AI and technology.

Week of March 16 to 22, 2026
Archive View latest review
Lead Story
Canada & Enterprise
CanadaRegulation

Canada Reverses TikTok Shutdown, Imposes New Data Sovereignty Conditions

After ordering TikTok Canada to wind down operations in November 2024, the Government of Canada reversed course following a federal court ruling, permitting TikTok to continue operating under new legally binding undertakings including security gateways, privacy enhancing technologies, and independent third party monitoring.

  • Security gateways and privacy enhancing technologies will control access to Canadian user data
  • An independent third party monitor will audit and verify data access controls continuously
  • TikTok must maintain a physical presence in Canada and invest in the cultural sector

Enterprise Impact: This decision establishes a precedent for how Canada will handle foreign owned technology platforms. The data sovereignty conditions may preview requirements that future regulation could extend to other cloud and AI platforms. Enterprises should monitor whether similar undertakings become standard for foreign technology vendors.

Source: BetaKit
CanadaEnterprise

Canada Invests $900 Million in Defence Industrial Strategy: Aerospace, Drones, and Quantum

The National Research Council received over $900 million under Canada’s Defence Industrial Strategy for aerospace defence capabilities, SME technology development, and quantum technologies. Key investments include a drone innovation hub in Ottawa and Montreal and $241 million through IRAP’s Defence Industry Assist initiative.

  • Over $500 million for next generation aerospace technologies, autonomous systems, and defence solutions
  • New drone innovation hub in Ottawa and Montreal
  • $241 million through NRC IRAP Defence Industry Assist for Canadian SMEs

Enterprise Impact: This is a massive signal for Canadian technology companies in aerospace, autonomous systems, quantum, and defence. SMEs with dual use technology should explore the NRC IRAP Defence Industry Assist stream. Suppliers on the Government of Canada AI Source List, including Unlock Solutions, are well positioned for adjacent AI and intelligent systems procurement.

Source: BNN Bloomberg
CanadaAIRegulation

Canada’s AI and Buy Canadian Strategies Set for 2026 Launch

AI Minister Evan Solomon confirmed that Canada’s national AI strategy and Buy Canadian policy will both launch in 2026, backed by $926 million over five years for sovereign public AI infrastructure. The Buy Canadian policy begins with defence and construction procurement.

  • $926 million over five years for sovereign AI infrastructure
  • Buy Canadian policy starts with defence and construction, expanding to full implementation by spring 2026
  • 145 pre qualified suppliers across 3 bands on the Government of Canada AI Source List

Enterprise Impact: The convergence of a national AI strategy with Buy Canadian procurement mandates creates a significant advantage for domestically based technology suppliers. Organizations already on the AI Source List are positioned to benefit from streamlined procurement. The sovereign compute focus signals government AI workloads will favour Canadian infrastructure.

Source: BetaKit
CanadaAI

Canada and Norway Commit to Sovereign Technology and AI Cooperation

Canada and Norway signed a joint statement committing to deepen collaboration on AI and digital technologies through the Sovereign Technology Alliance, building on Canada’s earlier agreement with Germany. The partnership focuses on reducing strategic technology dependencies and strengthening sovereign AI capacity across trusted democratic partners.

  • Norway exploring participation in the Sovereign Technology Alliance launched with Germany in February
  • Focus on sovereign AI capacity and reducing dependencies in strategic technology sectors
  • Builds on OECD and Global Partnership for Artificial Intelligence collaboration

Enterprise Impact: Canada’s growing network of bilateral AI agreements with democratic allies is establishing a framework for trusted AI collaboration. For Canadian technology companies, the expanding Sovereign Technology Alliance creates export and partnership opportunities with aligned nations. Organizations on the Government of Canada AI Source List may find new channels for international engagement.

Source: Government of Canada
AI Models & Platforms
AIEnterprise

Microsoft Launches Agent 365 and $99 Enterprise 7 Bundle, Adding Anthropic to Copilot

Microsoft announced Agent 365 ($15 per user per month) for governing AI agents across the enterprise, and Microsoft 365 Enterprise 7 ($99 per user per month) bundling Agent 365, Copilot, E5, Entra Suite, and advanced security. Copilot Cowork integrates Anthropic’s Claude alongside OpenAI models.

  • Agent 365 provides centralized governance for all AI agents regardless of underlying models
  • Enterprise 7 bundles the entire Microsoft AI and security portfolio at $99 per user per month
  • Copilot Cowork integrates Anthropic’s Claude alongside OpenAI models

Enterprise Impact: Agent 365 addresses the emerging governance gap as organizations deploy AI agents. The $99 Enterprise 7 bundle simplifies licensing for agentic AI at scale. Microsoft now offers model diversity (OpenAI plus Anthropic) as a feature. CIOs evaluating M365 Copilot should factor in multi model access when comparing platforms.

Source: Bloomberg
AIEnterprise

Mistral Launches Forge: Enterprise Custom Model Training Platform

Mistral announced Forge at NVIDIA GTC, a platform that lets enterprises build custom AI models trained on their own proprietary data using Mistral’s internal training recipes, data mixing strategies, and distributed computing optimizations. The company is on track to surpass $1 billion in ARR.

  • Forge packages Mistral’s internal training methodology for enterprise use
  • Also released Mistral Small 4 and Leanstral (open source code agent for formal verification)
  • Addresses AI project failure root cause: models trained on internet data rather than institutional knowledge

Enterprise Impact: Forge represents a significant shift toward enterprise model sovereignty. Organizations can train models on proprietary data without exposing it to third party platforms. This is particularly relevant for regulated industries where data cannot leave organizational boundaries. CIOs should evaluate whether custom model training delivers better outcomes than prompting general purpose models.

Source: TechCrunch
AIRegulation

Trump Administration Escalates Federal Ban on Anthropic AI Tools

The Trump administration intensified its campaign to remove Anthropic from all US government systems, with the Pentagon designating the company as a supply chain risk and developing replacement AI tools. The dispute centers on Anthropic’s refusal to allow its models to power mass domestic surveillance or fully autonomous weapons systems.

  • Pentagon ordered contractors to cease commercial activity with Anthropic within six months
  • Department of Justice stated Anthropic’s terms of service are “unacceptable to the Executive Branch”
  • Court filings revealed the Pentagon and Anthropic were nearly aligned before the executive order

Enterprise Impact: This dispute has direct implications for enterprise AI strategy. Organizations using Anthropic’s Claude in government adjacent work should assess contract exposure. The conflict illustrates that AI vendor selection now carries geopolitical risk: companies’ ethical positions on AI use cases can result in sudden loss of market access. Enterprise buyers should evaluate vendor diversification as a risk mitigation strategy.

Source: TechCrunch
Cloud & Infrastructure
EnterpriseAI

Microsoft Rolls Out AI Tools to Tackle $85 Billion Technical Debt Crisis

Microsoft launched autonomous AI agents designed to modernize legacy Java and .NET applications, targeting what McKinsey estimates is an $85 billion annual global cost from technical debt. The tools integrate with Azure Accelerate using agentic capabilities.

  • Technical debt costs the global economy over $85 billion annually
  • Organizations spend up to 40% of development resources maintaining legacy systems
  • Microsoft also acquired the team from Cove, a Sequoia backed AI collaboration platform

Enterprise Impact: This directly targets the number one barrier to enterprise AI adoption: legacy infrastructure. CIOs should evaluate these tools as part of modernization roadmaps, particularly for Java and .NET estates. The ROI case is compelling: redirecting even a fraction of the 40% maintenance burden toward innovation.

Source: VentureBeat
AIRegulation

Tesla Faces Escalated Federal Investigation into Full Self Driving Software

NHTSA escalated its investigation of Tesla’s Full Self Driving system to an “engineering analysis,” its highest level of scrutiny, covering 3.2 million vehicles. The probe focuses on nine crashes where the system failed to detect degraded camera visibility.

  • Probe upgraded to “engineering analysis,” the highest level before a potential recall order
  • Covers 3.2 million Tesla vehicles equipped with Full Self Driving
  • Systems failed to detect conditions that impaired camera visibility

Enterprise Impact: This investigation has direct implications for any enterprise deploying autonomous or AI assisted systems in safety critical environments. The core failure (systems not detecting when their own sensors are degraded) is a fundamental AI reliability challenge that extends to industrial automation and robotics. Enterprises should ensure their AI deployments include robust self monitoring and graceful degradation capabilities.

Source: Bloomberg
Enterprise Technology & Strategy
EnterpriseAI

Musk Announces Terafab: Joint Tesla, SpaceX, xAI Chip Manufacturing Facility

Elon Musk announced the Terafab project, a chip manufacturing facility in Austin, Texas jointly operated by Tesla and SpaceX for robotics, AI, and space data center chips. Tesla may tape out next generation AI6 chips in December, with Samsung securing a $16.5 billion manufacturing deal.

  • Terafab will be an advanced technology fab capable of making chips of any kind
  • Samsung secured $16.5 billion deal to produce Tesla’s AI6 chips
  • AI6 chips intended for self driving vehicles and humanoid robots

Enterprise Impact: The vertical integration of AI chip design and manufacturing signals a structural shift in the AI supply chain. Enterprises dependent on general purpose chips should monitor whether custom silicon advantages create competitive gaps. The shared chip facility across multiple companies could become a model for industrial AI infrastructure consolidation.

Source: Bloomberg
EnterpriseAI

Microsoft Rolls Back Copilot AI Bloat on Windows

Microsoft reversed course on some aggressive AI integration in Windows, pulling back Copilot features that users found intrusive. The move positions AI as a premium enterprise feature rather than a ubiquitous consumer overlay.

  • Microsoft scaling back mandatory Copilot integration in the consumer Windows experience
  • Shift positions AI as a premium enterprise feature
  • Comes alongside Agent 365 and Enterprise 7 as dedicated enterprise AI governance products

Enterprise Impact: This correction reveals a maturing approach to AI deployment: not everything needs AI. CIOs should apply the same lesson internally. Focus AI investment on workflows where it delivers measurable value rather than blanket deployment. Microsoft’s enterprise AI revenue strategy will center on the $99 Enterprise 7 bundle.

Source: TechCrunch
EnterpriseAI

Meta Deploys AI Content Enforcement Systems, Reducing Third Party Vendor Reliance

Meta announced advanced AI systems to handle content enforcement internally, replacing third party vendors. Also launching a Meta AI support assistant providing 24/7 user support globally across Facebook and Instagram.

  • Advanced AI systems replacing third party content moderation vendors
  • Meta AI support assistant rolling out globally for 24/7 user support
  • Part of broader in sourcing strategy alongside custom MTIA chip development

Enterprise Impact: Meta’s move to replace external vendors with internal AI systems is a leading indicator of how enterprises will rethink outsourcing. Functions previously outsourced (content moderation, customer support, compliance monitoring) are becoming candidates for AI driven in sourcing. Enterprises should evaluate which vendor relationships could be displaced by internal AI capabilities.

Source: TechCrunch
View all insights
Previous Reviews

Get the Weekly Briefing

AI & technology analysis delivered to your inbox every Monday. No noise, no hype — just what matters.