AI at a crossroads: How AI is reshaping infrastructure investment

The AI revolution is expected to profoundly impact the infrastructure space, creating significant opportunities for investment and enhancing value of existing infrastructure. Like any revolution, it will also bring about relevant challenges.

• AI is expected to drive growth in demand for key infrastructure, which we believe will create significant opportunity for investment.
• Implementation of AI is expected to power relevant improvements in financial and operational performance of infrastructure assets.
• A holistic risk management framework needs to be in place to identify and address emerging risks and ensure sustainable positive outcomes for all stakeholders.

Artificial Intelligence (AI) is a near-overnight business success, about 80 years in the making. While AI has been present and integrated in businesses for decades in one form or another, the most recent breakthroughs in generative AI have opened up an entire new world of possibilities. AI, like electricity and the internet, is a general-purpose technology that can impact many aspects of human civilization, and the infrastructure space is no exception.

The rapid adoption of generative AI is creating a surge in demand for key infrastructure to support deployment, thus creating an immense opportunity for investment in new projects. In addition, existing infrastructure assets are expected to derive relevant operational upside from AI adoption, thus increasing the value of these assets to society and their owners.

At the same time, AI poses a host of novel challenges that could have profound implications for human civilization. This article focuses more specifically on challenges that we believe will directly affect the infrastructure space—including increased cybersecurity risks, algorithmic issues (errors, biases), data issues (privacy, intellectual property), and labor challenges. Infrastructure investors and asset operators alike must ensure that AI systems are deployed thoughtfully and in consideration of all stakeholders.

The recent breakthroughs in AI were made possible by substantial increases in computing power, algorithm development, and the massive accumulation of data humankind has built up over decades.

AI experts agree that the momentum for AI development is unstoppable. While there is an element of hype at some levels of the “AI value chain", the total potential business value AI can create is so significant, that the fundamentals for continued strong growth in demand for data storage and computer processing power are undeniable.¹

This calls for massive investment in the complex infrastructure network that is required to run AI. This network is composed primarily of 1) data centers, 2) the electricity that is required to power them, and 3) the fiber network to connect them to users.

• Data centers – Global demand for storage and compute power in third party data centers has grown rapidly in the last decade, on the back of global corporate trends that include digitalization of processes and migration of data storage and compute power to the cloud. Global professional services company McKinsey estimates that global demand for data centers could reach 219 GW by 2030 (approximately quadrupling demand in 2023 and requiring investment of over $250 billion.¹)
• Electricity – According to the US Department of Energy, electricity demand from US data centers has grown dramatically in the last decade, climbing from 58 TWh in 2014 to 176 TWh in 2023, and is expected to climb to levels between 325 to 580 TWh by 2028². This increased demand is being driven by the electrification of many aspects of civilization, which we believe will create an enormous opportunity for investment in energy generation and grid enhancement.
• Fiber networks – Fiber network investment has experienced headwinds in recent years due to a combination of rising interest rates, increasing development costs and the saturation of the most economically attractive urban markets. The development of large-scale data centers that need to be connected to users—especially those that are being developed in secondary and emerging data center markets (further away from large urban centers)—are creating more attractive conditions for investment in new fiber network deployment.

Case study: Switch – data center platform³

IFM Investors made its first investment in a data center platform in 2022, whereby IFM partnered with DigitalBridge and acquired all outstanding shares of common stock in Switch, Inc. (NYSE: SWCH) in a take-private transaction with an enterprise value of ~$11 billion (as of 6 December 2022). At the time, Switch operated 16 top-tier data center facilities located across five strategically located campuses in the US,and has since opened three new facilities and has expansion plans for nearly 20 more in the next decade.

We believe Switch is not only a strategic asset providing the proverbial ‘picks and shovels’ to large customers with significant demand for data center services, but it also has advantages and distinct market positioning amongst data center providers. Switch was founded in 2000 and has built a tremendous business, team, and know-how that is valuable during a high-growth period that the business has now entered.

This is evidenced by Switch’s profitability, renewable energy utilization, large landbank within its operating campuses, and its dedicated construction and procurement teams, which can accelerate its ability to grow its platform. Further, we believe Switch’s proprietary data center designs and diversification in customers, regional markets, and physical data center locations, effectively mitigate downside risk.

IFM views Switch as a data center builder and operator finely-tuned by decades of experience that is uniquely positioned to capitalize on the many tailwinds for storage and computing.

Generative AI is expected to create sizeable economic benefits, given its potential to transform work and to accelerate technological development across industries. Global professional services company Accenture estimates that Generative AI can create up to $19.9 trillion in economic value by 2038, with a people-centric approach that maximizes productivity growth.⁴

Infrastructure assets are expected to benefit from this positive trend in multiple ways. First, IFM anticipates that the ensuing economic growth will bring increased demand for all types of infrastructure assets, including transportation, utilities and energy-related infrastructure, positively impacting the top line.

Second, a broad deployment of AI is enabling material improvements in efficiency through increases in productivity and optimization of operations across the value chain of infrastructure assets. Together with increased demand, these efficiency gains will have the potential to translate into enhanced returns for infrastructure investors.

Finally, infrastructure assets are positioned to achieve tangible improvement in the quality of services they provide, including safety and customer service, significantly enhancing the value of these assets to end users and society at large.

• Safety first – Despite being well into the 21st century, humanity still faces an unacceptable toll from accidents, with nearly 1.2 million transit-related fatalities⁵ and almost three million work-related deaths annually⁶ worldwide. The wide adoption of new technologies powered by AI, including autonomous vehicles, robotics, and human enhancement technologies, will enable infrastructure companies to dramatically reduce, and, hopefully, eventually eliminate such tragic events.
• Customer experience – Users of infrastructure assets can anticipate enhanced services driven by a broad range of AI-powered applications. For example, road users will enjoy lower rates of congestion due to improved traffic prediction and incident management technologies. Airport users will enjoy better experiences through improved safety and security technologies, such as biometric screening, while utility users are expected to enjoy increasingly agile and effective customer support interactions due to enhanced customer service technologies.
• Efficiency – Like most industries, infrastructure stands to benefit from AI solutions that boost labor productivity (see discussion below on AI’s impact on labor) and enhance the efficiency of both administrative and operational processes. In particular, infrastructure assets will benefit from improved asset maintenance processes and energy efficiency. All of these efficiency enhancements are expected to contribute to the bottom line and to the sustainability of infrastructure operations, thereby increasing risk adjusted returns for investors.
• Revenue optimization – Infrastructure assets will benefit from improved capabilities to predict usage volumes and reduce congestion. AI powered models will increasingly benefit from exponential growth in the deployment of sensors, and the increasing ability of AI to make sense of new and massive amounts of data. In turn, these developments will improve assets’ ability to stimulate volume, increase utilization, and assist in better timing capex expansion programs.
  

Case study: Manchester Airports Group⁷

At Manchester Airports Group (MAG), which encompasses Manchester, East Midlands and Stansted airports, IFM is driving transformation by harnessing the power of AI. MAG partners with service providers that help enhance the company’s digital and AI capability across all areas of the organization, allowing for operational optimizations and enhanced customer experience.

MAG’s AI models can predict how many passengers will be present at security during different times of the day. The forecasting module, which is benefiting from technological breakthroughs in AI algorithms, has now become accurate enough to drive meaningful operational benefits. With such accurate predictions, the manual effort required to produce end-to-end operational plans is significantly reduced, allowing enhanced planning, supply and resourcing, and ultimately better service.

The results of these AI deployments speak for themselves. Indeed, during a six-month span across MAG’s airports tracked from March to August 2024, 98% of passengers passed through security in less than 15 minutes, while at Manchester Airport seven million more passengers pass through security in less than five minutes than before the technology was used.

Further, MAG co-developed an innovative, industry-first solution, which combines flight data, weather conditions, and historical trends to more accurately calculate flight arrival and departure times. During its first months in operation, the model was 5-10% more accurate at predicting actual airfield movements versus the scheduled timing. If accuracy can be maintained and increased, the opportunity lies in maximizing aircraft movements during peak times.

We believe that MAG demonstrates the potential improvements AI can bring to existing operations, and IFM aims to apply similar strategies across our portfolio of airports.

While AI will have a broad and profound impact in the long run, we are focusing on its more immediate impact on infrastructure assets.

Similar to other industries, the infrastructure sector faces an intensification of existing challenges, such as cybersecurity threats and the impact of automation on the workforce. It must also contend with new risks, including algorithmic issues stemming from generative AI solutions. These challenges can have potentially harmful effects on the numerous stakeholders connected to infrastructure assets.

Compounding the challenge, the fast-moving evolution of the technology leads to a growing gap in the development of regulation that will be necessary to manage and mitigate these risks in a consistent and system-wide fashion.

This situation calls for a proactive approach by companies in developing robust frameworks that allow owners and managers to identify, understand, and address the emerging risks brought about by the technology. Below is a non-comprehensive list of challenges posed by AI, which are relevant to infrastructure assets.

• Cybersecurity – AI is arming bad actors with increasingly powerful tools that allow them to produce more sophisticated attacks, at a significantly lower cost. Generative AI generated code, deep fakes and automated phishing agents, are only some of the resources that are being used to increase the volume and effectiveness of well-known threats. In this context, infrastructure assets—most of which provide essential services to society—need to increase their efforts to ensure that the protections they have in place are adequate for the increased levels of risk.
• Algorithmic issues – Unlike traditional code that produces deterministic outputs based on the inputs it receives, generative AI models produce outputs that are based on probabilistic parameters. These parameters are set internally through exposing the model to vast amounts of data that allow “training” of the model. The implication is that these models, despite showing increasingly impressive human-like abilities, are not free of errors, biases and even hallucinations. They also potentially expose companies to data privacy and intellectual property issues. This means that any application of these algorithms needs to consider a broad set of risks in order to ensure that potential risk events are within the tolerance thresholds, and that appropriate mitigations are in place.
• Labor impact – Generative AI marks a major breakthrough by displaying abilities previously limited to humans, including the proficient use of natural language and creativity. This is enabling the technology to take on complex tasks such as coding, designing, painting, and even composing music. This has enormous implications for human labor, as the technology can perform a relevant portion of tasks that only humans could do until now. Consultancy Accenture estimates that over 40% of human working hours are in scope for automation or augmentation in the US.⁸

Generative AI is arguably one of the greatest advancements in human history, with the potential to reshape industries, improve lives, and solve complex challenges. Its ability to enhance decision-making and drive innovation makes it a powerful tool across sectors, not only for infrastructure.

To fully harness its benefits, it must be developed and implemented in ways that maximize its positive impact while proactively addressing potential risks, such as cybersecurity and misinformation. With thoughtful integration and responsible adoption, generative AI can unlock new efficiencies, fuel economic growth, and create investment opportunities that drive progress and improve quality of life across society.

Download the PDF

IFM Investors is a private markets manager with a focus on infrastructure and capabilities in unlisted and listed equity, private equity and debt. We’re based on four continents, with a global network that helps us uncover opportunities.

IFM was founded by pension funds. It’s still owned by pension funds today. Our clients are stewards of capital too, with the same duty to deliver value over decades. We share a multi-generational vision: to invest, protect and grow the savings of millions of people. We do it by investing, owning and financing the assets our clients and their beneficiaries rely on to live their lives.

_{1 McKinsey - AI power: Expanding data center capacity to meet growing demand
² Berkeley Lab – Supported by the US Department of Energy - 2024 United States Data Center Energy Usage Report
³ Case studies are provided for illustrative purposes only and should not be relied on to make an investment decision.
⁴ Accenture - Work, workforce, workers - Reinvented in the age of generative AI
⁵ World Health Organization – December 2023
⁶ International Labour Organization – November 2023
⁷ Case studies are provided for illustrative purposes only and should not be relied on to make an investment decision
⁸ Accenture - Work, workforce, workers - Reinvented in the age of generative AI}

_{Important Disclosures}

_{The following disclosure applies to this material and any information provided regarding the information contained in this material. By accepting this material, you agree to be bound by the following terms and conditions. The material does not constitute an offer, invitation, solicitation, or recommendation in relation to the subscription, purchase, or sale of securities in any jurisdiction and neither this material nor anything in it will form the basis of any contract or commitment. IFM Investors (defined as IFM Investors Pty Ltd and its affiliates) will have no liability, contingent or otherwise, to any user of this material or to third-parties, or any responsibility whatsoever, for the correctness, quality, accuracy, timeliness, pricing, reliability, performance, or completeness of the information in this material. In no event will IFM Investors be liable for any special, indirect, incidental, or consequential damages which may be incurred or experienced on account of a reader using or relying on the information in this material even if it has been advised of the possibility of such damages.}

_{Certain statements in this material may constitute “forward looking statements” or “forecasts”. Words such as “expects,” “anticipates,” “plans,” “believes,” “scheduled,” “estimates” and variations of these words and similar expressions are intended to identify forward-looking statements, which include but are not limited to projections of earnings, performance, and cash flows. These statements involve subjective judgement and analysis and reflect IFM Investors’ expectations and are subject to significant uncertainties, risks, and contingencies outside the control of IFM Investors which may cause actual results to vary materially from those expressed or implied by these forward-looking statements. All forward-looking statements speak only as of the date of this material or, in the case of any document incorporated by reference, the date of that document. All subsequent written and oral forward-looking statements attributable to IFM Investors or any person acting on its behalf are qualified by the cautionary statements in this section. Readers are cautioned not to rely on such forward-looking statements. The achievement of any or all goals of any investment that may be described in this material is not guaranteed.}

_{Past performance does not guarantee future results. The value of investments and the income derived from investments will fluctuate and can go down as well as up. A loss of principal may occur.}

_{This material may contain information provided by third parties for general reference or interest. While such third-party sources are believed to be reliable, IFM Investors does not assume any responsibility for the accuracy or completeness of such information.}

_{This material does not constitute investment, legal, accounting, regulatory, taxation or other advice and it does not consider your investment objectives or legal, accounting, regulatory, taxation or financial situation or particular needs. You are solely responsible for forming your own opinions and conclusions on such matters and for making your own independent assessment of the information in this material. Tax treatment depends on your individual circumstances and may be subject to change in the future.}

_{An infrastructure investment is subject to certain risks including but not limited to: the burdens of ownership of infrastructure; local, national and international economic conditions; the supply and demand for services from and access to infrastructure; the financial condition of users and suppliers of infrastructure assets; changes in interest rates and the availability of funds which may render the purchase, sale or refinancing of infrastructure assets difficult or impractical; changes in environmental and planning laws and regulations, and other governmental rules; environmental claims arising in respect of infrastructure acquired with undisclosed or unknown environmental problems or as to which inadequate reserves have been established; changes in energy prices; changes in fiscal and monetary policies; negative economic developments that depress travel; uninsured casualties; force majeure acts, terrorist events, under insured or uninsurable losses; and other factors beyond reasonable control.}

_{Australia Disclosure}

_{This material is provided to you on the basis that you warrant that you are a “wholesale client” or a “sophisticated investor” or a “professional investor” (each as defined in the Corporations Act 2001 (Cth)) to whom a product disclosure statement is not required to be given under Chapter 6D or Part 7.9 of the Corporations Act 2001 (Cth). IFM Investors Pty Ltd, ABN 67 107 247 727, AFS Licence No. 284404.}

_{Germany Disclosure}

_{This material is provided to you on the basis that you warrant that you are a professional client as defined in the relevant local implementation of the Markets in Financial Instruments Directive 2014/65/EU (“MiFID II”). This material is not intended for and should not be relied on by any other person. IFM Investors (Netherlands) B.V. shall have no liability, contingent or otherwise, to any user of this material or to third parties, or any responsibility whatsoever, for the correctness, quality, accuracy, timeliness, pricing, reliability, performance, or completeness of this material.}

_{Japan Disclosure}

_{This material is being provided to you by IFM Investors (Japan) Pty Ltd ("IFM Japan") (Financial Business Operator: Kanto Local Finance Bureau Register Number 2839, a member of the Type 2 Financial Instruments Firms Association). This material is intended only for use by a Qualified Institutional Investor (as defined in Article 10, Paragraph 1 of the Cabinet Office Ordinance Concerning Definitions Provided in Article 2 of the Financial Instruments and Exchange Act of Japan) and is not intended for anyone other than foregoing. This material is provided to you for informational purposes only and does not constitute the provision of investment advice. In addition, the information contained in this material does not constitute an offer to sell nor a solicitation of any offer to buy and is not intended to be, and should not be construed as, an offer to sell nor a solicitation of an offer to buy, any type of securities, financial products or other services offered by IFM Japan and its affiliates to any person in Japan to whom such offer or solicitation would be unlawful under the applicable laws, regulations and guidelines promulgated by the relevant Japanese governmental and regulatory authorities and in effect at the relevant time.}

_{Netherlands Disclosure}

_{This material is provided to you on the basis that you warrant that you are a Professional Investor (professionele belegger) within the meaning of Section 1:1 of the Dutch Financial Supervision Act (Wet op het financieel toezicht). This material is not intended for and should not be relied on by any other person. IFM Investors (Netherlands) B.V. shall have no liability, contingent or otherwise, to any user of this material or to third parties, or any responsibility whatsoever, for the correctness, quality, accuracy, timeliness, pricing, reliability, performance, or completeness of this material.}

_{United Kingdom Disclosure}

_{This material is provided to you on the basis that you warrant that you fall within one or more of the exemptions in the Financial Services and Markets Act 2000 (“FSMA”) [(Financial Promotion) Order 2005] [(Promotion of Collective Investment Schemes)(Exemptions) Order 2001, or are a Professional Client for the purposes of FCA rules] and as a consequence the restrictions on communication of “financial promotions” under FSMA and FCA rules do not apply to a communication made to you. IFM Investors (UK) Ltd shall have no liability, contingent or otherwise, to any user of this material or to third parties, or any responsibility whatsoever, for the correctness, quality, accuracy, timeliness, pricing, reliability, performance, or completeness of the information in this material.}

_{Switzerland Disclosure}

_{This Information is provided to you on the basis that you warrant you are (i) a professional client or an institutional client pursuant to the Swiss Federal Financial Services Act of 15 June 2018 ("FinSA") and (ii) a qualified investor pursuant the Swiss Federal Act on Collective Investment Schemes of 23 June 2006 ("CISA"), for each of (i) and (ii) excluding high-net-worth individuals or private investment structures established for such high-net worth individuals (without professional treasury operations) that have opted out of customer protection under the FinSA and that have elected to be treated as professional clients and qualified investors under the FinSA and the CISA, respectively.}

_{IFM-20FEB25-4221578}