Artificial Intelligence Q3 2025 Global Report - Hot topics, deal trends, and market outlook

Key Takeaways

• AI Moves from Experimentation to Integration: AI is now core to enterprise strategy with 65% of companies regularly utilizing generative AI (GenAI), doubling from 33% in 2023. This signals the shift in focus from AI pilots to structural adoption.
• Agentic AI Represents Growth Opportunity: Cowen projects enterprise spend on agentic AI will rise from less than $1 billion in 2024 to $51.5 billion by 2028, expanding at a ~150% Compound Annual Growth Rate (CAGR), as reasoning models and autonomous workflows enable enterprises to capture more value from AI integration.
• AI Infrastructure Powers the Economy: Massive funding requirements and outsized value opportunities presented by the AI buildout are fueling data infrastructure investments by Big Tech, hyperscalers, leading AI platforms, and private capital investors.

• Big Tech Investments: Leading tech companies are making massive investments in the buildout of data centers.
• Infrastructure Partnerships: AI platforms, chip providers, and datacenter providers are forming partnerships to scale faster.
• Public-Private Consortiums: Partnerships between the public sector and private investors are also forming to accelerate innovation and guide AI innovations to achieve national and societal benefits.
• Private Capital Investment: Private equity (PE) investment in AI remains highly concentrated in AI infrastructure due to stable projected demand growth and recurring revenues, consistent with PE’s picks and shovels approach.

• AI Deal Activity Surges: The pace of private capital investment in AI-focused companies is likely to remain elevated, as venture-backed startups make progress in commercializing innovative use cases and investors remain eager to capitalize on strong projected growth in the sector and optimism regarding AI’s transformative power.

Hot Topics

• High Adoption, Low Transformation: Despite $30 billion - $40 billion of enterprise investment in GenAI, 95% of companies have so far seen little to no Profit & Loss impact from AI, according to a recent MIT study. The 5% that extract value are embedding AI into workflows with memory and adaptation frameworks. Agentic AI could represent the next opportunity for enterprises to maximize value through greater automation and efficiency and reduced costs.

• Labor Market Facing Transitional Shifts: Goldman Sachs Research estimates that just 2.5% of U.S. employment would be at risk of displacement if current AI use cases were expanded across the economy. Looking ahead, innovation related to AI could displace 6-7% of U.S. jobs, though analysts expect the disruption to be temporary as AI-related innovation could also create new, higher-value employment opportunities. Occupations facing the highest risk of displacement by AI in the coming years include computer programmers, accountants and auditors, legal and administrative assistants, customer service representatives, telemarketers, proofreaders and copy editors, and credit analysts.

• Cyber Risks Intensify with AI: The volume and complexity of cyberattacks are escalating, with 81% of ransomware events in 2024 attributed to threat actors utilizing AI, according to an MIT study.
  • Weaponization of Agentic AI: AI models are being used to execute sophisticated cyberattacks.
  • Lower Barrier to Cybercrime: AI-enabled cyberattackers no longer need technical skills to develop ransomware.
  • Cybercriminals Embed AI Throughout Operations: Cyberattackers use AI for victim profiling and automated service delivery.
  • AI Utilized in All Stages of Fraud Operations: AI use has proliferated in fraud schemes including for data theft.

     

AI Deal Trends: Global Activity

STRATEGIC M&A

• Through Q3, 2025 strategic M&A involving AI-related targets has already surpassed all prior years on both a volume and value basis, with deal value up 294% YoY.
• Transaction activity involving AI targets is poised to stay elevated through the remainder of the year, as companies race to build or acquire AI capabilities.
• Amid the intensifying “fear-of-missing-out,” discipline and intentionality remain core to many acquirers’ M&A strategies, particularly in the context of organizations reporting high AI adoption rates but few transformational results.

PRIVATE EQUITY

• While PE firms are primarily investing in AI infrastructure, PE deals involving non-infrastructure AI targets have already reached a record level in 2025, with 265 deals through Q3, compared with just 161 through Q3 2024, representing an increase of 65% on a YoY basis.
• Rather than taking on the uncertainty of new AI platform investments, firms are enhancing the capabilities of existing portfolio companies through targeted add-on acquisitions. In Q3, 78% of PE deals with AI targets were add-ons, reflecting GPs’ preferences for creating value and managing risk in the AI space through buy-and-build strategies.

VENTURE CAPITAL

• While VC deal count for rounds involving AI targets is on pace to finish the year down 13% YoY, this is primarily a reflection of a slowdown in the broader VC market. Venture dollars invested in AI through Q3 2025 have already exceeded all previous years with AI minting billion-dollar startups faster than the dot-com era.
• AI continues to be a bright spot amid a slump in the broader VC market, with a growing bifurcation between the funding environment for AI vs. non-AI startups. AI-related investments accounted for 51% of global VC deal value, compared with 37% in full-year 2024 and 26% in 2023.
• A similar bifurcation is occurring by deal size for AI vs. non-AI startups, as AI unicorns are receiving a larger share of total venture dollars invested.

Notable AI Transactions: Venture Capital

• The VC market has experienced a pronounced increase in capital concentration, characterized by a widening gap between AI and non-AI funding rounds, as well as between unicorns and emerging startups.
  • Across all sectors, total VC deal count is on track to be the lowest in years, while total VC deal value is on pace to exceed 2024 and 2023, driven by investments in AI companies.

• Despite capital concentration, AI-enabled companies with unique applications or industry-specific use cases continue to raise large rounds, as demonstrated by Treeline Biosciences’ $600 Mn Series B.

Notable AI Transactions: Strategic M&A / PE

• Corporate repositioning is a major driver of M&A deals involving strategic buyers and AI targets, with companies shifting product mixes to capture higher-margin areas or adding AI solutions to enhance existing platforms.
  • Hewlett Packard’s $16 billion acquisition of Juniper Networks enables it to capitalize on the growing AI and hybrid cloud market opportunity by further establishing an AI-driven information technology portfolio.
• While PE investment in AI remains concentrated in infrastructure, and the vast majority of PE deals involving AI vertical products are add-ons (78% in Q3), PE investors retain an appetite for AI-enabled products with clear use cases.
  • TPG recently acquired Proficy, an AI-enabled software for industrial automation and efficiency, for $600 Mn.
• Technology companies, AI platforms, and AI infrastructure providers continue to invest in AI through acquisitions, partnerships and robust levels of capital spending to secure more compute and support AI deployment.
• CoreWeave’s $9.0 billion acquisition of Core Scientific is expected to enhance operational efficiency and expand data center capabilities.

The Infrastructure Buildout: Overview

$3 trillion - $4 trillion will be spent on AI infrastructure by the end of the decade, according to NVIDIA CEO Jensen Huang.

• AI developers and infrastructure providers are forging strategic alliances with perceived front-runners across the AI value chain and creating next-generation AI powerhouses, as evidenced by NVIDIA’s recent $100 billion partnership with OpenAI.

“Everything starts with compute. Compute infrastructure will be the basis for the economy of the future, and we will utilize what we’re building with NVIDIA to both create new AI breakthroughs and empower people and businesses with them at scale." — Sam Altman, Co-Founder and CEO of OpenAI (September 28, 2025)

Big Tech Investments

• Leading technology companies are making unprecedented investments in the buildout of their AI infrastructure.
• Microsoft committed $80 billion for data centers globally, the largest single AI infrastructure investment announced in 2025 to date.
• Google plans to spend $85 billion in 2025 to build out data center capacity and recently announced plans to invest $15 billion over five years to build an AI infrastructure hub in southern India.
• Meta plans to spend $600 billion on data centers and AI infrastructure through the end of 2028.

Infrastructure Partnerships

• AI platforms are partnering with chip providers like Broadcom, Marvell, Alchip, MediaTek, and AMD.
• In October, AMD and OpenAI announced a strategic partnership whereby AMD will provide compute power for OpenAI’s AI systems. Leading data center platforms have also established partnerships with NVIDIA to develop facilities and technology stacks that are AI-ready.
• NVIDIA, TSMC, and others formed a $500 billion manufacturing partnership, the largest AI hardware production commitment to date.

Public-Private Consortiums

• Partnerships are forming between the public sector and private AI developers as nations compete to develop the best technology.
• France, Italy, the UK, and Spain are collaborating with NVIDIA and regional cloud providers to deploy Blackwell AI infrastructure so they can build and run sovereign AI applications.
• The UAE launched Stargate with OpenAI and NVIDIA, leveraging state-owned G42 to construct a data center cluster.
• Saudi Arabia’s PIF and Google Cloud formed a partnership worth $10 billion to build an AI hub in Saudi Arabia.

Private Capital Investments

• Along with investments from AI platforms and sovereign investors, private capital firms are also providing critical capital to finance the AI buildout.
• Private equity is taking a “picks-and-shovels” approach to AI investment, opting to invest lower in the technology stack in AI infrastructure rather than in developers of AI applications that VC investors are targeting.
• BlackRock, Global Infrastructure Partners, MGX (Emirati-owned), and Microsoft launched the Global AI Infrastructure Investment Partnership to invest up to $100 billion in AI data centers.

The Infrastructure Buildout: Private Capital

PRIVATE EQUITY

• While venture investors are financing the capital needs of earlier-stage AI companies, PE firms are investing in companies that provide the back-end infrastructure to power AI platforms.
• BlackRock’s GIP has signed a deal to acquire Aligned Data Centers in a $40 billion transaction—one of the largest deals in 2025 and a display of PE confidence in AI infrastructure.
• While PE deal volume is projected to fall below 2024 by year end, deal value would already exceed 2024 with the inclusion of the announced $40 Bn Aligned Data Centers buyout (expected to close in H1 2026).
• Consistent with 2025 trends observed in the broader private capital market, capital deployed by PE investors is concentrated among the top cohort of managers who deployed the majority of capital into fewer target companies.
• PE investors regard data centers as stable, long-term investments that generate recurring revenues even amid macroeconomic volatility.
• Large asset managers with established infrastructure platforms are also capitalizing on data center tailwinds.
  • Brookfield Asset Management has already built 2,000 megawatts of data center capacity and expects the AI infrastructure buildout to be one of the largest capital formation cycles of this generation.

“A data center is a ‘Steady Eddie’ investment. You’re going to be able to model out a pretty predictable [internal rate of return] for that, and then you can make some more risk-forward bets in other places.” —Brad Haller, Senior Partner at West Monroe (June 5, 2025)

PRIVATE CREDIT

• Private credit is increasingly playing a role in financing AI infrastructure investments, especially in GPU-heavy buildouts.
  • In August, Meta arranged a $29 billion hybrid financing for its data center expansion with $26 billion in debt raised from private capital firm PIMCO.
  • Innovative financing models including GPU infrastructure-backed debt are emerging, with large private credit investors like BlackRock, JPMorgan, and The Carlyle Group issuing loans collateralized by GPUs.

VENTURE CAPITAL

• While VC investment dollars are highly concentrated in horizontal AI platforms and vertical applications, VC firms are looking to capitalize on AI tailwinds across the value chain and are also making investments in AI infrastructure providers.
  • In September, AI-native infrastructure provider Nscale secured $1.1. billion in the largest Series B round in European history, positioning the company to meet rapidly growing enterprise demand for GPUs and highlighting the role of venture capital in fueling the broader AI infrastructure buildout.

AI Market Outlook & Performance

MARKET TRENDS

• Growth Trajectory: Bloomberg expects generative AI (GenAI) to become a $1.3 trillion market by 2032, growing at a 42% CAGR, with the primary contributors to additional revenue growth being infrastructure services.

“We believe AI is providing a renaissance and macro tailwind to the broader technology sector with the potential to surpass the impact of the internet, cloud computing, or the mobile revolution.” —Jim Labe, Co-CEO and Co-Founder of TriplePoint Capital (June 5, 2025)

• AI Infrastructure: AI infrastructure investment is showing no signs of slowing down in the foreseeable future, with strong demand for computing power and intensifying competition among sovereign investors, AI platforms, technology companies, and private capital investors.

• Agentic AI: In the third quarter, adoption of agentic AI accelerated as enterprises moved beyond experimentation to strategic deployment. When combined with memory frameworks, and integrated effectively into workflows, agentic AI can help organizations overcome previous challenges to turning AI adoption into measurable, value-generating results.

  • Reasoning models now dominate agentic AI development and are projected to drive over 70% of applications by 2029 (from near zero in 2024), which is expected to fuel enterprise adoption.
  • By 2028, ~33% of enterprise software applications are expected to incorporate agentic AI, up from less than 1% in 2024, enabling automation of an estimated 15% of daily work decisions.

• Quantum AI: Recent advances in quantum computing, namely in quantum error correction, signal a strong long-term outlook for the convergence of quantum computing and AI, where quantum machine learning could eventually accelerate AI model training.

DEALMAKING OUTLOOK

• Venture Capital: Top investors are successfully raising and deploying capital into the leading AI horizontal platforms at unprecedented valuations, as well as startups with niche AI vertical applications. Bifurcation between leaders and emerging players in both the VC and AI markets is expected to persist until the exit environment materially improves. In the meantime, record funding rounds will continue to drive deal value as VCs compete to fund promising AI companies.
• Strategic M&A: The surge in strategic M&A volume involving AI targets is driven by the desire of incumbent firms to make transformative acquisitions and add high-growth offerings, and to enhance their current product mix through AI integration. Consistent with a positive outlook for acquisitions of AI targets by strategics, AI-focused strategic partnerships are also expected to become more prevalent particularly for companies investing in AI infrastructure.
• Private Equity: Buyouts of AI-related targets are expected to remain elevated relative to historical levels. However, drawn by predictable returns and rising demand for computational power, PE firms are expected to continue channeling capital into AI infrastructure. Improving exit conditions will eventually fuel participation from mid-market GPs.