Annuities and your portfolio at J.P. Morgan
How trust may drive a competitive edge in the age of AI
Will trust become the new competitive edge in the age of AI? The question is at the heart of today’s AI debate.
AI has dominated markets and investor attention, with U.S. large-cap tech stocks propelling the S&P 500 to record highs and delivering three consecutive years of 15%+ returns through the end of 2025. This initial phase of the AI trade was defined by tangible constraints and the capital deployed to address them – semiconductor supply, data center buildout, and securing sufficient power capacity. Early AI beneficiaries were often firms closest to these capital flows.
As the AI trade matures, attention is shifting to a harder-to-measure metric, the potential return on that capital investment. As we explore here, the next wave of AI winners may be not just the companies with the most powerful technology or strongest infrastructure, but those that people trust.
The word may be difficult to define precisely. But for economies and societies to function, a baseline level of trust must exist among countries, companies, customers and employees. As AI becomes embedded in decision-making, security and daily life, companies that build trust may unlock a competitive advantage. What does that mean in practice? Securing their infrastructure, gaining community acceptance and helping their workforce adapt.
In this article we explore how trust may drive a competitive edge and resilient investment value across three dimensions:
- security, systems and infrastructure
- communities and consumers
- workforce management and adaptability
We also focus on the potential investment implications across public and private markets.
Building trust: Security, systems, infrastructure
Trust in secure systems and infrastructure may create a durable advantage. As the world becomes more fragmented and AI’s capabilities accelerate, companies and countries are redefining what it means to be secure and consumers are moving toward systems they trust. National security, supply chain resilience, and safeguarding digital and physical infrastructure—these are top priorities for both policymakers and corporate CEOs.
Security threats are not new, but AI is raising the stakes. For example, cyberattacks have long presented a security threat, but AI introduces novel speed and capabilities for both attackers and defenders.
Two recent examples are instructive. In September 2025, AI firm Anthropic’s Claude Code was manipulated to launch cyberattacks across multiple countries with minimal human intervention.1 The threat has since escalated. Anthropic’s new Mythos model demonstrated an unprecedented ability to detect and exploit software vulnerabilities. Given these capabilities, Anthropic has yet to release the model publicly. Instead it has given select partners early access through its Project Glasswing to patch vulnerabilities before adversaries can exploit them.
OpenAI is taking a similar approach. Increasingly, businesses will be using AI to defend against AI.
Trust in cybersecurity defense
Critical infrastructure, such as healthcare, financial services and power grids, continues to be a top target of cyberattacks.2 CEOs globally ranked cyber risks as the greatest threat in 2026, tied with macroeconomic volatility.3 The high stakes are driving continued growth in cybersecurity spending globally. Forecasters project it will reach USD 230 billion this year and USD 320 billion by 2029.4 AI security M&A activity is projected to increase 400% in 2026.5
Highlighting the potency of the cybersecurity threat and AI’s impact, the U.S. Department of Defense has requested USD 15.1 billion for cyber activities in fiscal year 20266 and called for reduced guardrails in its use of AI models. More and more nations are reshoring digital infrastructure.7 To secure critical cloud systems, governments’ “sovereign cloud” spending is projected to reach USD 80 billion by 2026 and USD 111 billion by 2027, up from USD 59 billion in 2025.8
Power grids present a particular vulnerability, as seen in recent attacks on grids in Ukraine, Venezuela, and Iran. Clearly a major attack on a major economy’s grid could be very damaging. As far back as 2015, a joint report by Lloyds and the University of Cambridge Centre for Risk Studies estimated that the economic impact of a cyberattack on the U.S. power grid could potentially total USD 1 trillion.9
Investment implications
Where can investors find opportunities in the growing need for security? Cybersecurity firms and cyber insurance providers will likely benefit as demand for trusted security providers grows – especially as threats expand to smaller businesses.
In our view, cyber firms leveraging AI with deep customer integrations and partnerships will be able to build durable competitive moats. Over time, quantum computing may pose new risks to encryption, and companies that address these vulnerabilities could attract increased attention. Broadly, we think established companies with strong brands, resilient supply chains and proven secure technology systems may be increasingly valued by investors.
Google’s USD 32 billion purchase of cybersecurity firm Wiz in March 2026 underscores the sector’s momentum.
Global cybersecurity spending (USD billions)
Source: Gartner and J.P. Morgan estimates. Data as of June 20, 2025.
Building trust among communities and consumers
Community and consumer trust can impact the pace and scale of AI growth. Facing growing demand for AI, companies are racing to secure the computing power they need in new data centers. Up to USD 5.2 trillion in data center related capital investment could be deployed by 2030.10
But that projected spending may confront local resistance. Companies need to secure the trust of the local communities where data centers will be housed. Much like extractive industries such as mining or energy, data center developers are tethered to specific geographies—driven by the need to be close to customers, reliable power and reasonable cooling costs.11
Increasingly, though, their plans are met with skepticism or outright opposition as residents worry about rising electricity bills and growing strains on water supply. Those concerns will resonate as the subject of affordability animates public debate ahead of the November U.S. midterm elections. Developers must consider reputational and execution risks, especially in regions that are water-stressed or face regulatory uncertainty. Some forecasters project that up to 50% of 2026 data center projects may be delayed due to community opposition, grid challenges and skilled labor shortages.12
We think data center developers, particularly the five major hyperscalers (Microsoft, Meta, Oracle, Google and Amazon) will need to demonstrate they are trustworthy neighbors.
To earn community and customer trust, some developers are deploying advanced cooling systems and “bring your own generation” (BYOG) solutions to minimize water and power impacts on customers. Some are also pursuing bill-relief and ratepayer-focused initiatives alongside local utilities. Developers of data center projects that build strong relationships with local communities are well positioned to compete over the long term. And investors can find opportunities to identify companies that are building trust and those that are eroding it.
Growing protests sparked a wave of data center cancellations in 2025
Proposed data center projects cancelled after sustained local protests, by quarter
Source: Heatmap Pro. Data as of January 12, 2026.
Building trust in educational, media and medical AI
Of course, the issue of AI trust extends beyond infrastructure. AI tools are shaping education, youth safety and development, and decision-making across a wide range of activities. Trust in the tools depends partly on how they are trained and rolled out. Mounting lawsuits over child safety, data privacy and intellectual property, such as Anthropic’s USD 1.5 billion settlement for using copyrighted books for training,13 suggest that trust may be quite fragile.
So-called deepfakes (AI-enabled sham video, audio or images) synthetic content (AI-altered text, video or audio) and algorithm-driven echo chambers on social media are eroding trust in what users see, hear and read. In education, AI models offer tailored teacher lesson plans that can be useful. But excessive reliance on AI tools by students may compromise their problem-solving skills and cognitive development.14
Trust in medicine is also shifting. Some 230 million ChatGPT users – about 30% of total users – now ask health questions weekly, including over a million facing serious issues like suicidal intent or psychosis.15 Many observers worry this reflects misplaced trust in AI medical guidance and a declining reliance on actual medical professionals, with some users becoming emotionally dependent on the AI models.
Meanwhile, the workplace presents its own set of trust issues, as we discuss in the following section.
In legal, medical, and financial services AI models can streamline workflows. But the models may raise unresolved questions about liability, quality and malpractice. As AI tools become more capable, will companies be seen as negligent if AI isn’t used? If a doctor or lawyer relies on AI and a mistake occurs, who is responsible?
Investment implications
Ultimately, the quality and safety of AI model outcomes often depend on human critical judgment. Companies enabling identity verification, content labeling and AI safety – including those helping navigate legal liability - may be well-positioned as adoption increases and scrutiny intensifies. Companies that build trust with communities and consumers could gain a competitive edge.
Building trust in workforce adaptability
Trust in workforce management and adaptability can lead to sustained returns on AI investment.
In the coming months and years, companies and investors will likely focus increasingly on whether firms are generating a sufficient return on their AI investment.
As AI technology becomes more integrated into daily workplace operations, trust between employees and employers may be a key differentiator in delivering that investment return. We think companies that can achieve genuine employee buy-in, rather than superficial adoption, will be better able to deliver durable investment gains.
Productivity gains and a changing job market
Return on investment depends in part on productivity gains. We are beginning to see early signs that AI is boosting productivity in the U.S. economy. What does this mean for workers? Some are concerned that AI technology is improving so rapidly that labor markets will not have time to adjust. But we believe that AI will increase economic productivity and corporate profits at a pace that allows for a manageable reconfiguration of labor markets.
The effects of AI adoption will vary widely across sectors and individual companies. In the near term, sectors with a high share of automatable tasks (such as entry-level technology roles) will likely continue to see the first wave of impacts.16
Computer science graduates now face higher unemployment rates than those with degrees in art history or philosophy17— though this trend may not follow a linear trajectory as industries adapt to AI. The buildout of AI infrastructure is driving demand for harder-to-automate roles like electricians, while creating entirely new roles such as AI specialists. The result: Some jobs are in great demand while people with the "wrong" skills struggle to find work. Aging populations in many developed economies may exacerbate strains as labor markets adapt to AI.
Meanwhile, high-profile departures by key AI researchers suggest that even top talent has growing concerns about the technology's direction. Once again, trust is a key part of the conversation.
Investment implications
What are the investment implications of building (or not building) trust in the workplace?
Across industries and sectors, we think companies and investors will increasingly value reliable AI solutions in which human expertise and judgment are critical components.
We believe that companies that address employee anxiety around job security and professional value, maintain transparency and invest in talent pipelines and development are more likely to achieve lasting productivity gains. This may be especially relevant as AI usage costs rise and some technology companies plan workforce reductions.18 Investors should monitor whether firms are realizing cost efficiencies through layoffs or delivering revenue growth with steady employment.
Companies supporting the AI buildout via skilled trades may see new demand and greater resilience to AI disruption. Firms less correlated to AI disruption - including the so-called HALO (heavy asset, low obsolescence) assets such as utilities, telecoms and power generation networks - may also see growing demand. Their businesses may be increasingly attractive to investors looking to bolster portfolio diversification and reduce concentration in the AI trade.
Conclusion
These are early days, but AI looks to be ushering in a profound transformation of our economy and society. Trust—in security, communities and workforce adaptability—may be a defining factor in harnessing the power and managing the risks of AI.
Ultimately, humans are resilient. Trust in that resilience and our collective ability to adapt will be essential as the AI era unfolds.
-
Anthropic, “Disrupting the first reported AI-orchestrated cyber espionage campaign,” November 13, 2025.
-
IBM Security, Cost of Data Breach Report 2025. FBI Internet Crime Report 2024. J.P. Morgan Strategic Research.
-
PwC 2026 Global CEO Survey.
-
Gartner and J.P. Morgan estimates.
-
Momentum Cyber. “Cybersecurity market review”, April 2026
-
Congress.gov, FY2026 Department of Defense Cyber Budget Request
-
RBC, “The Great Recalibration: Forces of Exponential Change,” November 17, 2025. Some 70 countries require that data be stored and processed within their borders.
-
Gartner, “Worldwide Sovereign Cloud IaaS Spending,” date.
-
Llyod’s and the University of Cambridge estimate of a severe but plausible cyberattack on the US power grid.
-
McKinsey & Company, “The cost of compute: A $7 trillion race to scale data centers,” date. The spend to 2030 could add 125 GW of new capacity.
-
NOVVA Data Centers, “Does Data Center Location Really Matter? The Hidden Geography of Data,” date.
-
Sightline Climate, Q1 Data Center outlook. February 2026.
-
NPR, “Anthropic settles with authors in first-of-its-kind AI copyright infringement lawsuit.” Date.
-
Also called automation bias. Giuseppe Romeo andDaniela Conti, “Exploring automation bias in human–AI collaboration: a review and implications for explainable AI,” AI & Society 41,July 2025.
-
OpenAI, Strengthening ChatGPT responses in sensitive conversations
-
Stanford Digital Economy Lab, Canaries in the Coal Mine? For example, employment in AI-exposed roles dropped around 16% for early-career workers (age 22-25) since late 2022.
Anthropic, Labor market impacts of AI: A new measure and early evidence.
-
Federal Reserve Bank of New York. Labor Market Outcomes of College Graduates by Major in 2024. Released February 2026.
-
Harvard Business Review, Why AI adoption stalls according to industry data. Microsoft plants first-ever voluntary employee buyout for up to 7% of U.S. Workforce. Meta cutting 10% of its workforce (CNBC).
IMPORTANT INFORMATION
In the age of AI, trust—not just technology and infrastructure—may determine who leads, who lags and who lasts. Here’s what you need to know.
Daniel Rourke
Head of Sustainability Research & Integration, Private Bank
Published May 5, 2026
you may also like
Experience the full possibility of your wealth
We can help you navigate a complex financial landscape. Reach out today to learn how.
Contact us
