The Future of AI in Investments: Predictions for 2026 and Beyond

Executive Summary

NEW YORK, 2026 — The investment industry has crossed a threshold. What began as experimental copilots inside research desks and compliance teams is now embedded in production workflows managing trillions in assets. BlackRock's Aladdin platform oversees roughly $25 trillion; Goldman Sachs has rolled its AI assistant out to all 46,500 employees; Morgan Stanley reports a 98% adoption rate among financial-advisor teams. Yet the ROI picture is bifurcated: McKinsey estimates AI could restructure 25–40% of an asset manager's cost base, while PwC's 2026 CEO survey found 56% of chief executives got "nothing out of" their AI investments. This briefing sets out our core thesis — that 2026 is the year agentic AI moves from assistive to autonomous — and offers six evidence-anchored predictions for enterprise decision-makers.

Key Takeaways

• McKinsey estimates AI, gen AI and agentic AI could be equivalent to 25–40% of an asset manager's cost base, concentrated in distribution, investment processes, compliance and software development.
• BCG projects agentic systems can free 35–50% of client-coverage capacity by automating onboarding, reporting and routine requests.
• Adoption is now near-universal at bellwether firms: Morgan Stanley reports 98% FA-team adoption; Goldman deployed its GS AI Platform to all 46,500 employees.
• The ROI gap is real: PwC found 56% of CEOs saw no measurable return, and only ~29% of executives in IBM's data can confidently measure AI ROI.
• Capital markets themselves are being reshaped — Morgan Stanley has led or co-led some $65 billion in corporate bond deals for data centers or other AI investments since October 2025 — more than any other large US bank — according to data compiled by Bloomberg.
• The next phase is agentic: JPMorgan is targeting enterprise-wide AI-agent deployment by 2026 against a technology budget near $19.8 billion.

Market Analysis: The Economics Behind the Predictions

The central tension in the 2026 investment-AI market is a productivity paradox. McKinsey notes that technology spending has surged at an 8.9% CAGR across North America and Europe over five years, yet cost as a share of AUM has stayed "relatively flat at the industry level." The promise of agentic AI is that it finally converts that spend into structural efficiency rather than incremental tooling. McKinsey identifies portfolio-manager copilot agents and code copilots as the highest-impact, easiest-to-implement use cases, projecting productivity lifts of 25–40%.

The table below synthesises verified figures from the leading deployments and analyst studies referenced in this briefing. Figures are drawn from company press releases, McKinsey, BCG, PwC and Bloomberg reporting.

The dispersion between the top and bottom of this table is the story. Financial services leads all verticals on gen-AI returns precisely because it concentrates high-volume, rules-based knowledge work with the data infrastructure to support it. But the PwC and IBM data confirm that measurement discipline — not model access — separates winners from laggards.

Prediction 1: Agentic AI Moves From Copilot to Autonomous Operator

The defining shift of 2026 is agentic AI — systems that plan and execute multi-step tasks with limited human intervention. The signal is unambiguous. On 11 July 2025, Goldman Sachs said it was piloting an autonomous AI software engineer (Cognition's Devin) alongside its nearly 12,000 human developers — starting with hundreds of instances and potentially scaling to thousands, according to CIO Marco Argenti — with the firm projecting productivity gains of three to four times over previous tools. JPMorgan plans to deploy more powerful, longer-running AI agents in 2026 across knowledge-work and client-facing roles, chief analytics officer Derek Waldron told CNBC.

Related: Helsing $18B Defence AI Round 2026: Dragoneer, Lightspeed Lead $1.2B Raise

"lasting agentic advantage primarily requires changes to the operating model, talent, and processes that embed AI into how work gets done, not by data and technology alone." The prediction: by end-2026, agentic workflows will be standard in software development, compliance monitoring and client reporting at tier-one firms — but the differentiator will be operating-model redesign, not model selection.

Prediction 2: The ROI Gap Narrows for Disciplined Adopters — and Widens for Everyone Else

The most important nuance in the 2026 data is the ROI bifurcation. PwC's 2026 Global CEO Survey of 4,454 CEOs across 95 countries found 56% reported getting "nothing out of" their AI investments, and only 12% said AI had both grown revenues and reduced costs simultaneously. IBM's Think Circle work indicates only about 29% of executives can confidently measure AI ROI today.

For deeper context, see our Investments analysis: "Jimini Health, M13 & Zetta Target AI Behavioral Health Expansion in 2026".

Yet the firms with measurement discipline show hard numbers. Goldman developers report a 20% productivity increase and a 15% reduction in post-release bugs. Morgan Stanley advisors using its Debrief tool save roughly 30 minutes of administrative work per meeting, and executives attributed record performance — including almost $64 billion in net new assets in Q3 2024 and 100,000 new clients — partly to AI-driven efficiency and prospecting. The prediction: over 12–24 months, the ROI gap will not close uniformly. It will widen between firms that instrument their deployments and those that fund pilots without measurement frameworks. For a broader view of how enterprise AI economics are being measured, compare the trajectory of adjacent markets such as cybersecurity spending and risk.

Prediction 3: In-House Builds Displace Vendor Dependence at the Largest Firms

A quieter but structurally significant trend: the largest managers are building rather than buying. When Vanguard announced its Expert Insights portfolio-analysis tool on 9 April 2026 — currently in pilot with select advisors and slated for wider availability later in 2026 — it disclosed the tool was "developed in-house by Vanguard developers using an undisclosed foundational LLM." The demand driver was concrete — Vanguard's annual portfolio-analysis engagements with advisors quadrupled over six years. Meanwhile, BlackRock's Aladdin continues to serve as a platform layer for others: its October 2025 AI "Auto Commentary" feature launched with Morgan Stanley as first client, and its December 2025 AWS partnership will bring Aladdin to secure cloud infrastructure, with Amazon Treasury among the first adopters and general availability expected in H2 2026. The prediction: expect a two-tier market — mega-managers building proprietary agents on foundation models, while mid-market firms consume platform capabilities via Aladdin-style ecosystems and cloud providers.

Additional coverage: Secfix Targets European Compliance Automation Growth with $12M Raise in ...

Competitive Landscape

The competitive field spans incumbent asset managers, investment banks, platform providers and cloud infrastructure players. The table below maps the leading named deployments verified in this briefing.

Notably, AI is also reshaping capital-markets revenue itself. Bloomberg reckons Morgan Stanley has led or co-led some $65 billion in corporate bond deals for data centers and AI investments since October 2025 — a reminder that the sector both deploys and finances the AI boom. The scale of that financing appetite is echoed in the anticipated OpenAI IPO targeting a $1T debut.

Related: Eighteen48 Partners €175M Fund 2026: Mid-Market PE Strategy Explained

Practical Business Implications

For enterprise decision-makers, three actions follow directly from the evidence. First, instrument before you scale: the PwC and IBM data show measurement — not access — is the binding constraint on ROI. Establish baseline metrics (developer productivity, per-meeting admin time, compliance cycle time) before rollout. Second, prioritise the McKinsey-identified high-impact, low-complexity use cases: portfolio-manager copilots, code copilots, and compliance automation. Third, treat operating-model redesign as the deliverable, per BCG — embedding agents into workflows rather than bolting them onto existing processes. Firms in capital-intensive adjacent sectors face similar sequencing challenges, from aerospace decarbonisation targets to pharma's real-world data problem, where technology promise must survive measurement discipline.

Forward Outlook: 2026 and Beyond

Our overarching prediction is that 2026 marks the transition from AI-as-tool to AI-as-operator in investment management. The near-universal adoption figures at Goldman, Morgan Stanley and JPMorgan indicate the experimentation phase is over. What remains contested is value capture. McKinsey's 25–40% cost-base opportunity and BCG's 35–50% capacity gain are ceilings, not floors — realised only by firms that redesign operating models. Over the next 12–24 months, expect three durable shifts: proprietary agent stacks at mega-managers, platform consolidation around Aladdin-style ecosystems and cloud providers, and a widening performance gap between measurement-disciplined adopters and pilot-fatigued laggards. The technology question is largely settled; the management question is not.

For deeper context, see our Cyber Security analysis: "USTR Maintains China Tech Tariffs as BIS Tightens Cyber Export Controls".

Frequently Asked Questions

How much of an asset manager's costs could AI realistically address?

McKinsey estimates the combined impact of AI, generative AI and agentic AI could be equivalent to 25–40% of an average asset manager's cost base, with value concentrated in distribution, investment processes, compliance automation and software development. This is a potential ceiling, and realising it depends heavily on operating-model change rather than technology spend alone.

Why do so many firms report no return on AI investment?

PwC's 2026 Global CEO Survey found 56% of 4,454 CEOs got "nothing out of" their AI investments, and IBM data suggests only about 29% of executives can confidently measure AI ROI. The primary constraint is measurement discipline: firms that establish baseline metrics before deployment — as Goldman and Morgan Stanley did — show quantifiable gains, while those funding pilots without instrumentation struggle to demonstrate value.

What is agentic AI and why does it matter for investments in 2026?

Agentic AI refers to systems that plan and execute multi-step tasks autonomously rather than merely assisting a human. It matters because firms like Goldman Sachs (autonomous software-engineer agents) and JPMorgan (enterprise-wide agent rollout) are moving from copilots to operators. BCG stresses that lasting advantage requires embedding agents into operating models, talent and processes — not just adopting the technology.

Are the largest firms building or buying their AI?

A two-tier market is emerging. Mega-managers increasingly build in-house: Vanguard's Expert Insights, announced in April 2026 and initially in pilot with select advisors, was developed by Vanguard developers on an undisclosed foundation LLM. Simultaneously, platform providers like BlackRock's Aladdin serve mid-market firms, with its AWS partnership (December 2025) extending cloud-hosted AI capabilities to enterprise clients from H2 2026.

How is AI affecting investment banking revenue, not just operations?

AI is reshaping capital-markets revenue directly. According to Bloomberg, Morgan Stanley has led or co-led roughly $65 billion in corporate bond deals for data centers and AI-related investments since October 2025 — meaning the sector both deploys AI internally and finances the broader AI infrastructure buildout externally.

Sources include company disclosures, regulatory filings, analyst reports, and industry briefings.

Related Coverage

• More Investments Analysis
• Latest Technology News

Analysis based on company announcements, investor disclosures, regulatory filings, Reuters, Bloomberg, Financial Times, CNBC, SEC documentation, and publicly available market data as of publication.