Allianz, Pictet or BNP – Which AI Investment Strategy is yours?

November 7, 2025

Executive Summary

Three distinct types of AI investment strategies have emerged in European wealth management:

Type I: Humans invest in AI companies (thematic exposure)
Type II: AI invests automatically (under human supervision)
Type III: Humans invest with AI support (hybrid approach)

Each approach carries fundamentally different operational, regulatory, and competitive implications. We cover three prominent examples from Allianz, Pictet and BNP and help you identify which approach aligns with your institution’s capabilities and competitive positioning.

Introduction

As a wealth management professional in late 2025, you face a fundamental strategic question: which type of AI investment strategy do you follow – and is it the right approach for long-term competitiveness?

The distinction goes far beyond technology selection. When Allianz Global Investors structures a fund to capture the AI economy, when Pictet Asset Management delegates stock selection to deep learning models, and when BNP Paribas Wealth Management deploys AI to optimize client portfolios – each institution is making a different bet on how AI creates value in wealth management.

These are not parallel experiments. They represent three distinct business models, each with its own operational requirements, regulatory obligations, and competitive dynamics. Understanding which path fits your institution starts with understanding what each approach actually means in practice.

Type I: The Thematic Investor – Allianz Global Investors

The Approach: Humans invest in AI companies through traditional active management.

Allianz Global Investors’ artificial intelligence strategy represents one of Europe’s largest AI-themed investment vehicles. The fund allocates at least 70% of its portfolio to companies whose business operations are tied to AI development and deployment.

The investment universe spans the AI value chain: semiconductor manufacturers providing the hardware foundation, cloud infrastructure providers hosting AI systems, software developers creating AI applications, and companies implementing AI-driven automation in their operations.

What this means for you: The portfolio management process follows conventional active equity principles. Analysts conduct fundamental research on companies within the AI sector. Portfolio managers make allocation decisions based on company prospects, competitive positioning, and valuation. The “AI” component defines the investment universe, not the investment process.

This represents the majority of what the market labels “AI funds.” They provide sector exposure to the AI economy without introducing algorithmic decision-making into portfolio management itself.

Strategic implications: For wealth managers, this category requires no change to existing investment processes. It addresses client demand for AI exposure through familiar active equity frameworks. The value proposition is access to professionally managed exposure to AI-beneficiary companies – selecting which semiconductor manufacturer or cloud provider will outperform, which AI application companies warrant investment, and how to balance exposure across the AI value chain.

Type II: The Algorithmic Investor – Pictet Asset Management

The Approach: AI invests automatically (under human supervision) through systematic stock selection.

Pictet Asset Management launched its AI-driven global equities strategy in March 2024. Proprietary deep learning models serve as the primary driver of stock selection across global equity markets, analyzing company fundamentals, analyst sentiment, price patterns, and market activity to predict individual stock alpha.

Structured as a UCITS fund and registered across multiple European jurisdictions, this represents institutional-grade implementation of AI as portfolio manager rather than portfolio tool.

What this means for you: The AI model makes autonomous recommendations on which securities to hold. Human oversight focuses on risk management, portfolio constraints, and model governance rather than individual security selection. The system processes data volumes and identifies patterns beyond human analytical capacity.

This approach addresses a specific use case: systematic stock selection at scale. The value proposition is not human judgment about which stocks to buy, but computational power to identify pricing inefficiencies and alpha sources across thousands of securities simultaneously.

Strategic implications: This category represents a fundamental shift in value creation. The competitive advantage lies in model sophistication, data quality, and computational infrastructure rather than analyst insight or portfolio manager judgment. For institutions evaluating whether to develop or partner for AI-driven strategies, this demonstrates that regulatory approval and operational implementation are achievable. However, the technology infrastructure requirements, data science talent needs, and model governance frameworks represent substantial investments.

Type III: The Hybrid Investor – BNP Paribas Wealth Management

The Approach: Humans invest with AI support – combining human judgment with computational power.

BNP Paribas Wealth Management’s approach positions AI as decision support rather than decision maker. The institution’s algorithmic system processes client risk profiles, ESG preferences, time horizons, and current market conditions to generate optimal asset allocation proposals.

The system considers what the bank describes as “billions of possible combinations” to suggest portfolio weightings across asset classes. Yet the institution explicitly maintains that “the role of the advisor remains central” and that AI “does not diminish their role, but on the contrary enhances it.”

What this means for you: Investment advisors review, modify, and approve all recommendations before implementation. The AI provides computational optimization – handling the mathematical complexity of balancing multiple client constraints and market inputs – while humans provide judgment, client relationship management, and ultimate decision authority.

This augmentation model addresses a different challenge than autonomous AI: it enhances human judgment by providing computational power for optimization while preserving the advisor relationship and fiduciary responsibility with the client.

Strategic implications: For many wealth managers, this represents the most immediately actionable approach. It leverages AI capabilities without requiring fundamental restructuring of investment processes or client relationships. The value proposition combines human judgment with computational scale – advisors maintain client relationships and exercise professional judgment, while AI handles optimization complexity and processes data volumes beyond human capacity.

The operational requirements are more modest than full AI-driven strategies: integration with existing systems, training advisors to interpret and contextualize AI recommendations, and establishing governance around when to follow or override AI suggestions.

The Regulatory Dimension

These three approaches exist within the EU AI Act framework, which creates specific obligations depending on how AI systems are classified and deployed.

For Type I (Thematic) investors: Conventional UCITS rules apply. The “AI” component is a sector classification, not an operational AI deployment. No specific AI-related regulatory obligations arise from the investment strategy itself.

For Type II (Algorithmic) investors: Questions around algorithmic transparency, model governance, and risk management frameworks become material. Investment firms must address system documentation requirements, demonstrate appropriate human oversight, and establish processes for model validation and performance monitoring. The regulatory obligations reflect the systematic nature of AI-driven decision-making under human supervision.

For Type III (Augmentation) investors: The regulatory questions center on advisor responsibility and client disclosure. When AI generates recommendations that advisors typically follow, what constitutes adequate human oversight? How should firms document instances where advisors override AI suggestions? What must be disclosed to clients about the role of AI in their portfolio management?

The strategic and compliance dimensions of AI adoption have become inseparable considerations. The path you choose determines not just your technology roadmap, but your regulatory obligations and governance requirements.

Looking Forward

The European wealth management industry in November 2025 has moved beyond whether AI matters to which AI strategy you pursue. The three approaches – investing in AI, letting AI invest automatically (under supervision), or investing with AI – represent distinct strategic choices with different operational, regulatory, and competitive implications.

For Type I investors, the path is familiar: conventional portfolio allocation decisions within a thematic framework.

For Type II investors, the requirements are substantial: proprietary technology development or sophisticated partnerships with AI specialists.

For Type III – the hybrid approach – the opportunity lies in combining the best of both worlds. This is where human judgment, client relationships, and fiduciary responsibility meet computational scale and optimization power. It’s also where the implementation challenges are most nuanced: how to integrate AI tools effectively, how to establish appropriate governance, and how to navigate the regulatory framework while maintaining competitive advantage.

Take action now

The question is no longer whether to adopt AI, but which approach aligns with your institution’s capabilities, client base, and strategic positioning – and how to implement it effectively.

The institutions defining best practices in AI deployment today are establishing the competitive reference points for tomorrow. The window for deliberate positioning narrows as competitors deploy AI at scale and client expectations rise.

Should you wish to discuss your specific situation and implementation approach, please do not hesitate to reach out.

Sources:

About the Author: Dr. Andreas K. Janoschek specializes in AI applications for European Asset & Wealth Management. Based in Geneva, he helps industry professionals stay ahead of competition by securely advancing with AI.

This newsletter aims to inform and does not constitute investment or legal advice. Always consult with qualified professionals for specific circumstances.