Aperture Intelligence
All briefsSubscribe

Artificial Intelligence · February 22, 2026 · 21 articles

AI Governance Fractures as Trade Wars and Platform Shifts Reshape Global Order

Executive Summary

The global AI governance consensus fractured this week as 88 nations signed a toothless declaration in New Delhi while individual countries raced ahead with binding frameworks. India's AI Impact Summit produced a declaration heavy on democratization rhetoric and open-source endorsements but devoid of enforceable safety commitments — a stark departure from the safety-first framing of the 2023 UK summit. Italy, meanwhile, became the first EU member state to enforce a comprehensive AI workplace law with real penalties, and Canada deepened bilateral AI partnerships with India and Australia. The US trade architecture underwent a constitutional reset, with cascading consequences for the global economy. After the Supreme Court struck down Trump's IEEPA-based tariffs, the White House pivoted to Section 122 of the Trade Act of 1974, imposing a 15% global tariff for 150 days. Singapore's leadership acknowledged the country must prepare for a permanently altered trade environment, even as its Budget 2026 committed a record S$37 billion to research and elevated AI to a prime-ministerial priority. AI is simultaneously rewriting the infrastructure of commerce, knowledge discovery, and human expression. LinkedIn's B2B traffic collapsed 60% as AI search eliminated clicks; Google embedded shopping ads into AI Mode for 75 million daily users; OpenAI outlined autonomous advertising that could bypass agencies entirely. These shifts signal not just a platform transition but a fundamental restructuring of how human attention, commercial value, and information credibility are mediated — a transformation whose implications extend well beyond business cycles into how civilizations organize knowledge and allocate trust.

Key Takeaways

  • 01Global AI safety consensus collapses as nations prioritize access over guardrails: The New Delhi declaration signed by 88 countries contains zero binding safety measures, relying entirely on voluntary or industry-led commitments. This represents a decisive shift from the 2023 UK summit's safety-first orientation toward India's 'democratization' framing. Over the next decade, the absence of enforceable global AI safety norms means governance will fragment into competing regional regimes — the EU's binding approach, America's market-led model, and the Global South's access-first priority — creating regulatory arbitrage opportunities and systemic risks that compound as AI capabilities accelerate.
  • 02Constitutional checks force US trade policy into temporary legal frameworks: The Supreme Court's rejection of IEEPA-based tariffs forced Trump to invoke Section 122, which caps tariffs at 15% for 150 days maximum. This legal constraint introduces a hard deadline and ceiling that the previous framework lacked. The pattern of executive overreach followed by judicial correction followed by legal workaround suggests persistent trade uncertainty that will outlast any single policy — businesses must architect supply chains for permanent volatility rather than waiting for resolution.
  • 03Singapore bets S$37 billion that research investment can offset geopolitical fragmentation: The 32% increase in RIE2030 funding, a National AI Council chaired by the PM, and a dedicated AI park represent Singapore's most concentrated strategic investment in a generation. GDP growth is projected to slow to 2-4% from 4.8%, yet Singapore is accelerating investment rather than retrenching. This counter-cyclical posture — investing more aggressively precisely when the global environment deteriorates — offers a template for how small, open economies can convert uncertainty into structural advantage over multi-decade horizons.
  • 04AI search engines destroy traditional web traffic while creating new influence economies: LinkedIn's 60% B2B traffic decline and ChatGPT's low referral rates despite billions of daily prompts demonstrate that AI interfaces absorb rather than distribute attention. Google's 75 million daily AI Mode users and new in-conversation shopping ads show the monetization model adapting in real time. For the Anthropocene epoch, this represents a phase transition in how human knowledge is accessed — from navigating to destinations toward receiving synthesized answers, with profound implications for institutional credibility, epistemic diversity, and the economic models that fund original knowledge creation.
  • 05Italy establishes first binding AI employment framework with real enforcement teeth: Fines of up to €1,500 per employee, mandatory plain-language explanations of AI decisions, and a dedicated Ministry oversight committee make Italy's framework the most concrete AI workplace regulation in Europe. Combined with the EU AI Act's 'high-risk' classification for employment AI systems, this creates a compliance baseline that will likely propagate across member states. Over five to ten years, the employer obligation to explain AI decisions — not just disclose them — could fundamentally reshape how organizations design and deploy workforce automation.
  • 06OpenAI's autonomous advertising vision threatens to disintermediate entire agency models: OpenAI's monetization head described a system where businesses prompt ChatGPT to create, manage, and optimize campaigns conversationally — autonomously testing bids, allocating budgets, and refining strategy. Initial tests target US free-tier users. If AI can commoditize campaign execution, competitive advantage shifts to firms with proprietary data, cultural insight, and strategic judgment that machines cannot replicate — accelerating a bifurcation between high-value strategic work and automated commodity execution across all knowledge-intensive industries.
  • 07Startup ecosystems reveal that institutional quality matters more than market size: Singapore's 23.85 unicorns per 10 million people versus Nigeria's 0.08 demonstrates a 300x gap in startup density driven by institutional factors — regulatory clarity, capital access, and international connectivity. Nordic countries' strong top-ten showing confirms that social safety nets enable rather than inhibit entrepreneurship. Over generational timescales, this data suggests that the countries investing in institutional infrastructure today — not just technology — will capture disproportionate shares of the AI-driven economy's value creation.

Action Items

  • →[Immediate] Monitor the 150-day countdown on US Section 122 tariffs beginning February 24, and map supply chain exposure across affected trade corridors to identify hedging strategies before the July expiration deadline forces new policy pivots. (Addresses: Global trade resilience)
  • →[This Week] Review your organization's AI search visibility strategy in light of LinkedIn's 60% B2B traffic decline and Microsoft's three-stage brand surfacing framework, prioritizing structured data, entity authority, and contextual relevance over traditional keyword-based SEO investments. (Addresses: Digital strategy adaptation)
  • →[This Month] Assess compliance readiness against Italy's AI workplace law and the EU AI Act's high-risk classification for employment AI systems, benchmarking current disclosure practices and human oversight mechanisms against the plain-language explanation requirements that are now enforceable. (Addresses: AI regulatory compliance)
  • →[This Quarter] Engage with Singapore's RIE2030 ecosystem and Growth Capital Workgroup opportunities, evaluating whether your organization's research, innovation, or scaling activities align with the four funded domains — manufacturing, health, sustainability, and smart nation — to capture S$37 billion in strategic investment flows. (Addresses: Strategic positioning in Asia)
  • →[This Quarter] Convene a cross-functional working group to evaluate the implications of OpenAI's autonomous advertising model and Google's AI Mode commerce integration on your organization's customer acquisition costs, agency relationships, and competitive positioning in AI-mediated discovery channels. (Addresses: Future of customer acquisition)

Sources

𝕏Share on XShare on LinkedIn

Generate your own personalized briefings on the topics you choose. Multi-source synthesis, role-specific analysis, action items.

Sign up — free during beta