Markets Weigh AI Slowdown & Guardrails: Marc Franklin

Manulife Investment Management's Marc Franklin said the recent AI-stock pullback reflects uncertainty over regulatory guardrails and the pace of development, rather than damage to AI's structural growth outlook. He warned that changing regulation and open-source quantum tools could intensify competition and pressure the margins of incumbent high-margin technology suppliers.

Analysis

The relevant question is not whether AI spending persists, but where regulation shifts the profit pool. Rules that impose model-audit, provenance, privacy, and compute-reporting requirements favor hyperscalers with compliance teams, enterprise distribution, and proprietary data; they disadvantage smaller application vendors whose valuation assumes rapid, lightly regulated deployment. A slower deployment cadence can also defer SaaS seat expansion while leaving infrastructure demand relatively resilient, creating a near-term preference for cash-generative platforms over high-duration AI software.

Open-source quantum tooling is a longer-dated competitive issue rather than an earnings catalyst. If it lowers experimentation costs, it may commoditize portions of specialized optimization and security software, while increasing demand for cloud compute, data-management, and post-quantum cybersecurity migration. The non-obvious loser is legacy encryption-dependent enterprise infrastructure: procurement cycles for cryptographic agility can lengthen before revenues materialize, pressuring vendors unable to bundle security upgrades into broader platforms.

Positioning risk remains asymmetric over the next 1-3 months: AI beneficiaries with the highest expectations are vulnerable if regulatory headlines delay monetization or cap regional product availability, even without reducing multi-year capex. The structural thesis is falsified only by a broad hyperscaler capex reset, enterprise AI workload cancellation, or regulation that materially restricts commercial model access rather than raising compliance costs. Absent those signals, regulatory uncertainty is more likely to widen dispersion than invalidate AI exposure over 6-18 months.

Key Decisions for Investors

• Favor a quality AI barbell for the next 1-3 months: long MSFT and GOOGL versus a basket short of high-multiple AI application software via IGV. The thesis is that compliance and distribution costs consolidate enterprise AI economics; exit the pair if hyperscaler AI capex guidance is cut or IGV materially outgrows cloud-platform bookings.
• Add PANW or CRWD on regulatory-driven weakness as a 6-18 month post-quantum/security-compliance proxy, but size modestly until disclosed enterprise demand for cryptographic migration emerges. Risk is that quantum-security budgets remain a non-funded mandate and spending is displaced by cloud-security consolidation.
• Do not chase a broad AI-beta rebound through SMH until the next hyperscaler earnings cycle confirms capex durability. Use a break above the pre-pullback high only with evidence that accelerator lead times and cloud backlog remain firm; a broad capex-guidance reduction would argue for reducing semiconductor exposure.
• Monitor EU AI Act implementation milestones and U.S. federal compute-reporting actions as volatility catalysts. A regime that mandates audits but preserves deployment supports MSFT/GOOGL; restrictions on model access or material liability standards would warrant cutting AI-platform beta and increasing downside hedges through QQQ puts.