10 Key Truths About AI According to Morgan Stanley. What Investors Need to Know in 2026?

MarGib October 15, 2026

🌐 🇵🇱 Polski · 🇬🇧 EN

In its latest report from July 2026, Morgan Stanley presents 10 fundamental truths about artificial intelligence that every investor should know. Which sectors, companies, and trends will shape the AI market in the coming years? Let’s examine the data, forecasts, and risks.

Infographic showing stock charts of AI-related companies and market forecasts for 2026. — Visualization of market data regarding artificial intelligence in 2026.

Why AI is not a bubble, but a long-term innovation cycle

Morgan Stanley Investment Management (MSIM), in a report published on July 2, 2026, states unequivocally: artificial intelligence is not a passing trend, but a profound transformation comparable to the internet revolution. According to analysts, spending on AI solutions by S&P 500 companies grew by approximately 30% year-over-year in 2025, confirming the long-term nature of this phenomenon. Unlike the dot-com bubble at the turn of the century, AI is already generating real revenue and productivity in key sectors of the economy.

Key indicators confirming this thesis include:

Growth in AI revenue among tech companies – e.g., Microsoft Azure recorded a 45% increase in Q1 2026 (source: Microsoft Earnings).
Adoption of generative AI in enterprises – 70% of Fortune 500 companies are using these technologies in 2026 (McKinsey, June 2026).
Infrastructure investments – data centers and semiconductors are seeing record growth; for example, TSMC forecasts a 25% revenue increase in 2026 due to chip production for AI.

MSIM emphasizes that AI is not a homogeneous market, but an ecosystem consisting of many layers – from infrastructure to end-user applications. This makes investments in this technology have both short- and long-term potential.

AI Infrastructure: semiconductors, cloud, and data centers

According to Morgan Stanley, the foundation of artificial intelligence development is infrastructure. Without advanced semiconductors, efficient data centers, and scalable cloud solutions, further progress in AI would be impossible. The report identifies three key segments of this market:

1. Semiconductors: NVIDIA, AMD, and TSMC at the forefront

NVIDIA remains the undisputed leader in the production of graphics processing units (GPUs) dedicated to AI. In May 2026, the company's market capitalization exceeded $4 trillion, making it the most valuable technology company in the world. However, MSIM points to growing competition:

AMD – forecasts a 50% growth in AI revenue in 2026 thanks to new Instinct MI300 chips.
TSMC – as the primary chip manufacturer for NVIDIA and AMD, expects a 25% revenue increase in 2026.
Intel – despite delays, is introducing Gaudi 3 chips to the market, which are intended to compete with NVIDIA in the AI segment.

It is worth noting that China is intensively developing its own semiconductor technologies. In May 2026, Huawei introduced the Ascend 910B chip, which is intended to compete with NVIDIA's solutions. This could change market dynamics in the coming years.

2. Cloud computing: Microsoft Azure and Amazon AWS dominate

Cloud solutions are essential for training and deploying AI models. According to Synergy Research Group data for Q1 2026, Microsoft Azure and Amazon AWS together control approximately 60% of the market. MSIM points to several trends in this segment:

Microsoft Azure – AI revenue growth of 45% in Q1 2026, driven by integration with Office 365 and Dynamics 365 services.
Amazon AWS – is developing services like Bedrock, which allow companies to build their own AI models.
Google Cloud – despite a smaller market share, is recording 35% AI revenue growth in 2026 thanks to solutions like Vertex AI.

MSIM emphasizes that cloud computing is not just about platforms, but also an ecosystem of tools and services supporting AI development. An example is Microsoft Fabric, introduced in 2025, which integrates data, analytics, and AI into a single platform.

3. Data centers: energy and logistics as new challenges

AI development is associated with a massive demand for energy. According to a report by the International Energy Agency (IEA) from March 2026, AI data centers are already consuming approximately 2% of global electricity. This presents new challenges for the industry:

Sustainability – companies like Google and Microsoft are investing in renewable energy sources to offset the increase in consumption.
Energy infrastructure – energy companies, e.g., NextEra Energy, are seeing increased demand for power, which translates into their financial results.
Data center location – countries with access to cheap energy, such as Canada or the Nordic countries, are becoming attractive locations for new investments.

MSIM notes that data centers are not just costs, but also an opportunity for investors. Companies specializing in building and managing infrastructure, such as Equinix or Digital Realty, may see significant growth in the coming years.

AI Models: open-source vs. closed – who is winning?

One of the most important topics covered in the Morgan Stanley report is the rivalry between open-source and closed models. In 2026, both categories are developing dynamically, but they have different advantages and disadvantages.

Closed models: GPT-5 and Gemini 2.0

Closed models, developed by large corporations, continue to dominate in commercial applications. MSIM points to several key launches in 2026:

GPT-5 from OpenAI – introduced in March 2026, offers multimodal support (text, image, audio, video) and better performance in tasks requiring reasoning.
Gemini 2.0 from Google – launched in April 2026, integrated with Google Workspace, which allows for the automation of many business processes.
Claude 4 from Anthropic – focused on safety and ethics, gaining popularity in the financial and medical sectors.

The advantage of closed models is their high quality, technical support, and integration with large corporate ecosystems. The downside, however, is high licensing costs and limited flexibility.

Open-source models: Llama 3.1 and Stable Diffusion 3

Open-source models are gaining importance, especially among small and medium-sized enterprises and developers. MSIM highlights several key events in 2026:

Llama 3.1 from Meta – released in April 2026, reached 100 million active users in June 2026. It is free and competitive in quality with GPT-4.
Stable Diffusion 3 from Stability AI – launched in February 2026, offers 4K image generation and is widely used in the creative industry.
Mistral AI – a French start-up that gained popularity thanks to high-quality open-source models.

The advantage of open-source models is their accessibility and flexibility. Companies can adapt them to their needs without having to pay for licenses. The downside, however, is the lack of official support and potential issues with security and regulatory compliance.

MSIM predicts that in the long term, open-source models may gain importance, especially in sectors where costs are a key factor. However, in commercial applications requiring high quality, closed models will continue to dominate.

Edge AI: artificial intelligence on edge devices

Another important trend in 2026 is the development of Edge AI, which involves processing data directly on end devices such as smartphones, tablets, or IoT devices. Morgan Stanley points to several key benefits of this approach:

Speed – processing data locally eliminates latency associated with sending it to the cloud.
Privacy – data does not leave the device, which increases security and regulatory compliance.
Energy efficiency – local processing consumes less energy than transmitting data to the cloud.

MSIM lists several key players in this segment:

Apple – the A18 Pro chip, introduced in the iPhone 16 in September 2025, offers a 32-core Neural Engine, allowing for advanced AI processing directly on the device. You can read more about the new features in the iPhone 18 Pro in our article.
Qualcomm – the Snapdragon 8 Gen 4 chip, launched in January 2026, contains dedicated AI accelerators that allow for running models locally.
NVIDIA – the Jetson Thor platform, unveiled in 2025, enables running AI models with 120 billion parameters on edge devices. You can find more on this topic in our post.

Edge AI is of particular importance in sectors such as medicine, automotive (autonomous vehicles), and industry. MSIM predicts that by 2030, over 50% of IoT devices will be equipped with local AI processing capabilities.

AI Regulations: how do laws affect the market?

The development of artificial intelligence does not take place in a regulatory vacuum. In 2026, regulations introduced by the European Union, the United States, and other regions are of key importance. In its report, Morgan Stanley emphasizes that regulations can both hinder and stimulate innovation in AI.

European Union: AI Act fully implemented

The EU AI Act, which fully enters into force in August 2026, introduces a number of requirements for companies developing and deploying AI. Key assumptions include:

AI system classification – the regulations divide AI systems into four risk categories: unacceptable, high, limited, and minimal.
Ban on certain applications – e.g., biometrics in public places or social scoring systems.
Mandatory audits – high-risk systems must undergo regular audits and compliance assessments.

MSIM notes that the AI Act may slow down innovation in Europe, but at the same time, it increases security and trust in AI technology. Companies that comply with the regulations may gain a competitive advantage in the global market.

United States: AI Executive Order and new regulations

In the USA, the key document is the AI Executive Order from October 2023, which was updated in March 2026. New regulations introduce, among other things:

Mandatory AI model testing before commercialization.
Safety standards for AI systems used in critical infrastructure.
International cooperation regarding AI regulation, especially with the EU and G7 countries.

MSIM emphasizes that regulations in the USA are less restrictive than in the EU, which may favor innovation. At the same time, the lack of uniform federal-level regulations can lead to legal uncertainty.

China: technological and regulatory rivalry

China is intensively developing its own AI regulations, which aim to both control the technology and support domestic companies. Key events in 2026:

New regulations on generative AI – introduced in January 2026, require, among other things, the registration of AI models and their compliance with "socialist values."
US chip export restrictions – the update to regulations in March 2026 limits China's access to advanced semiconductors.
Development of domestic technologies – companies like Huawei and Alibaba are developing their own AI models, e.g., Pangu 4.0, which premiered in May 2026.

MSIM assesses that China is catching up in AI but still lags behind the USA in terms of infrastructure and talent. However, regulations may help in building a competitive AI ecosystem.

Investment risks in AI: what should investors know?

Investments in artificial intelligence carry a number of risks, which Morgan Stanley discusses in detail in its report. Here are the most important ones:

1. Regulations and legal compliance

As mentioned, regulations such as the EU AI Act or the American AI Executive Order can significantly impact the development of the AI market. MSIM points to several key threats:

Compliance costs – companies must invest in audits, certifications, and adapting their systems to regulations.
Application restrictions – some AI systems may be banned or restricted, which will affect their profitability.
Legal uncertainty – dynamically changing regulations can lead to unpredictable market changes.

2. Competition and the market

The AI market is extremely competitive, which creates both opportunities and threats for investors:

US-China rivalry – China is investing heavily in AI, which may reduce the dominance of American companies.
Open-source models – free alternatives, such as Llama 3.1, lower barriers to entry for new players and may reduce the profits of large corporations.
AI-washing – some companies misuse the term "AI" for marketing purposes, which can mislead investors.

3. Technological limitations

Despite rapid development, AI still faces a number of technological limitations:

Energy consumption – AI data centers consume huge amounts of energy, which can lead to availability and cost issues.
Talent shortage – there is a global shortage of approximately 1 million AI specialists (source: Kaggle Survey 2026), which may slow down technology development.
Ethics and safety issues – AI models can generate false information, discriminate against users, or be used for criminal purposes.

4. Bubble risk

MSIM warns against overly optimistic valuations of AI-related companies. Many companies are seeing record growth, but not all have real business fundamentals. Key indicators to pay attention to:

AI revenue as a % of total revenue – e.g., NVIDIA generates 85% of its revenue from AI, while other companies may have a much smaller share.
R&D spending – market leaders such as Meta or Alphabet allocate 15-20% of their revenue to research and development.
Number of patents – IBM, Microsoft, and Google are leaders in terms of the number of AI-related patents.

AI market forecasts to 2030: what do analysts say?

In its report, Morgan Stanley presents forecasts regarding AI market growth through 2030. Here are the key figures and assumptions:

AI market value in 2030: $1.8 trillion (CAGR ~37% between 2026–2030).
Key segments:
- AI software and services: $1.2 trillion.
- Infrastructure (semiconductors, data centers): $600 billion.
Regions:
- USA: 50% of the market (dominance in infrastructure and software).
- China: 30% of the market (rapid growth in commercial applications).
- EU and the rest of the world: 20% of the market.

For comparison, other institutions present different forecasts:

McKinsey: $2.6 trillion by 2030 (CAGR 40%).
IDC: $1.5 trillion by 2030 (CAGR 30%).
Gartner: $2.0 trillion by 2030 (CAGR 35%).

MSIM emphasizes that these forecasts are based on several key assumptions:

Continued innovation – development of new AI models, such as multimodal systems or Edge AI.
Enterprise adoption – an increase in the number of companies using AI in their business processes.
Infrastructure development – investments in data centers, semiconductors, and cloud computing.
Regulations – regulations that will favor innovation rather than hinder it.

AI investment strategies: how to profit from artificial intelligence?

In its report, Morgan Stanley presents several investment strategies that can help investors leverage the potential of AI. Here are the most important ones:

1. Individual stock investments

For investors who want direct exposure to AI development, MSIM recommends several key companies:

NVIDIA (NVDA) – leader in GPU production for AI, with a market capitalization exceeding $4 trillion in 2026.
Microsoft (MSFT) – dominance in cloud computing (Azure) and integration of AI with products like Office 365.
Meta (META) – development of open-source models (Llama) and AI applications in advertising.
TSMC (TSM) – key chip manufacturer for NVIDIA and AMD.
Alphabet (GOOGL) – development of Google Cloud and AI models such as Gemini.

2. ETFs

For investors who prefer diversified exposure to AI, MSIM recommends several ETFs:

Global X Robotics & AI ETF (BOTZ) – exposure to robotics and AI, with growth of ~50% in 2025.
iShares Robotics and Artificial Intelligence ETF (IRBO) – global AI-related companies.
Roundhill Generative AI & Technology ETF (CHAT) – focused on generative AI, launched in January 2026.
ARK Autonomous Technology & Robotics ETF (ARKQ) – investments in autonomous technologies and AI.

3. Hedge funds and venture capital

For institutional and advanced investors, MSIM points to several strategies:

Hedge funds:
- Renaissance Technologies – AI-based quant strategies.
- Two Sigma – investments in AI and big data companies.
Venture Capital:
- Sequoia Capital – investments in AI start-ups, such as Mistral AI.
- Andreessen Horowitz – support for companies developing AI models and applications.

4. Corporate bonds

For investors seeking stable income, MSIM recommends corporate bonds of AI-related companies:

Microsoft 2030 – yield of ~4.5% in 2026.
Alphabet 2029 – yield of ~4.2% in 2026.

5. Private Equity

For long-term investors, MSIM points to private equity funds investing in private AI companies:

Blackstone AI Fund – investments in AI start-ups and mature AI companies.
KKR Technology Growth Fund – support for technology companies, including AI.

Summary: 10 key truths about AI according to Morgan Stanley

In its July 2026 report, Morgan Stanley presents 10 fundamental truths about artificial intelligence that every investor should consider:

AI is not a bubble, but a long-term innovation cycle – comparable to the internet revolution.
Infrastructure is key – semiconductors, cloud, and data centers drive AI development.
Open-source models are gaining importance – but closed models still dominate in commercial applications.
Edge AI is the future – processing data on edge devices eliminates latency and increases privacy.
Regulations may hinder innovation – the EU AI Act and US regulations introduce new requirements for companies.
AI is changing the job market – by 2030, it may automate ~30% of tasks in the service sector. You can read more about this in our article.
Ethics and responsibility are crucial – standards such as ISO/IEC 42001 are gaining importance.
US-China rivalry is shaping the market – China is catching up, but the USA maintains an advantage.
AI is revolutionizing the financial sector – algorithmic asset management and fraud detection.
AI market forecasts are optimistic – the market value could reach $1.8 trillion by 2030.

Artificial intelligence is one of the most important technologies of our time, with the potential to change almost every aspect of the economy and daily life. Investors who understand its dynamics, trends, and risks will be able to effectively leverage its potential in the coming years.