Navigating the AI Era: The Convergence of Generative AI and Regulatory Realignment in Investment Research

The investment  research industry is undergoing an unprecedented structural transformation. Driven by the rapid, ubiquitous integration of artificial intelligence (AI) across the value chain and a synchronised alignment of global regulation, the economics of investment insights are being fundamentally rewritten. As market participants grapple with flat nominal budgets, escalating expenses, and evolving valuation frameworks, the operational relationships between buy-side investors, sell-side brokers, and corporate issuers are being reoriented towards quality and investment process integration.

 

The Macro Dynamics of Modern Research Valuation

The institutional research ecosystem is currently defined by four interconnected structural changes:

• Velocity of Technological Adoption: Artificial intelligence is transforming  content creation, distribution and integration into workflow  at a pace that exceeds historical trends, such as the move to Cloud or HTML content
• Structural Budget Pressure: Aggregate research spending has stagnated in real terms, while additional assets such as alternative data, AI tokens and API feeds are competing for share within the same pool  
• Disruption of Traditional Attribution: The transition from static document consumption to AI-ready content  has broken legacy tracking metrics, complicating how the sellside services the buyside and protects IP 
• Global Regulatory Alignment: For the first time in nearly a decade, international regulatory frameworks are pointing in the same direction, clearing a behavioural pathway for a widespread return to commission-funded research models.

 

Research Budget Bifurcation and the Operational Squeeze

Global research budgets are flat, forcing an operational divergence across the industry. Institutional spending is actively shifting away from traditional screen-based consumption toward direct, LLM-ready content ingestion. This behavioural shift creates a widening divide between technology-forward market participants and legacy providers.

The sellside faces margin pressure. Procuring enterprise-grade AI architecture requires heavy capital expenditure, top-tier talent remains expensive, yet the buyside signals flat wallet allocations. Compounding this pressure is the emergence of “tokenflation” risks. As early technology vendor subsidies diminish, the operational costs of running advanced Model Context Protocol (MCP) or Large Language Model (LLM) tokens are rapidly converging with traditional human headcount expenses.

Industry scrutiny regarding these massive technology investments is expected to peak within the next six to twelve months. During this window, research providers must definitively demonstrate that their augmented platforms generate measurable alpha. Scaled, data-rich, talent-heavy sell-side firms with highly differentiated content will capture a larger share of a shrinking operational pie, while providers of commoditised, routine analysis will face aggressive budget cuts.

 

Buyside Mandate: Elevating Analytical Alpha

Investment managers have moved past utilising AI purely for efficiency. Instead, the strategic focus has shifted toward scaling advanced tools directly within proven investment frameworks to uncover unique alpha.

Buyside analysts increasingly deploy specialised agents for pre-meeting preparation. These tools autonomously parse historical notes, macro news flows, earnings transcripts, expert network reports, and broker research to generate highly tailored lines of questioning for corporate management interactions.

Crucially, this expanded capacity does not prompt investment teams to broaden their watchlists. It allows them to scale intensive analytical work that was previously cost-prohibitive, such as running multi-variable scenario models and cross-comparing massive external data points. This shifts professional hours away from routine data processing toward higher-order investment thesis generation. To facilitate this workflow, traditional broker budgets and data budgets are expected to converge within the next 24 months, giving investment teams the structural flexibility to reallocate capital dynamically as data feeds directly into internal LLMs.

 

The Sellside Pivot: Monetisation in a Machine-First World

For sellside firms, generative AI represents both an existential threat to routine informational output and an unprecedented commercial opportunity to monetise proprietary intellectual property at scale. Sellside research analysts now devote roughly one-third of their standard working hours to managing and reviewing AI workflows.

Routine touchpoints between buyside clients and sellside analysts have declined in frequency due to the use of LLMs to answer questions often directed to sales and research professionals. However, this shift has significantly increased the value, depth, and concentration of the human interactions that do take place.

Investment leaders note that cultivating deep financial intuition and a rigorous, detail-oriented work ethic in junior analysts is far more challenging when automated platforms handle the foundational, iterative tasks that historically built those professional skills.

Concurrently, a “golden age” of research monetisation is emerging for firms with rich historical archives, as these datasets carry significant new value as machine training inputs. Distribution frameworks built on agent-readable, semantically tagged data can now provide precise attribution down to individual data fields. This technological evolution has sparked intense intellectual property debates. Sellside institutions argue that legacy commission relationships do not automatically grant buyside firms the right to ingest complete research libraries into proprietary LLMs, meaning separate commercial valuations and licensing agreements are required.

 

Corporate Access: The Resilient High-Touch Anchor

While written analysis and raw data face rapid commoditisation, physical and high-touch corporate access remains exceptionally resilient. Industrial volume data shows corporate access activity rising 10% to 15% year-on-year, led by a 10% growth in non-deal roadshows (NDRs). This represents a decisive return to pre-pandemic baselines as institutional investors prioritise highly concentrated meetings over mass-participation industry conferences.

Conferences are still the  primary platform and account for 75% of corporate access slots, but concentrated engagements – such as an hour-long meeting between executive leadership and a small group of high-conviction investors – are valued far more. Technology is primarily optimising the logistics of these events through advanced scheduling analytics, allocation tools, and AI-driven transcription services that deliver text outputs within minutes of a presentation closing.

Importantly, corporate issuers are exerting greater control over their investor relations. Recognising that the industry has historically over-indexed on buyside meeting demand, corporates are actively prioritising engagements that help them strategically curate their long-term shareholder register. This dynamic has driven increased reliance on expert networks, which offer unvarnished private channel insights alongside traditional management access.

 

Re-Engineering the Valuation and Feedback Loop

The mechanics of evaluating research value are shifting to combined  qualitative human feedback with  automated consumption metrics.

• Granular Evaluation Demands: For example, adding analyst-level granularity in broker votes rather than aggregated firm-level scores. Firms will typically direct high-value human resources toward more transparent clients.
• The Interaction Metric Problem: Current evaluation frameworks still over-index on the raw volume of client interactions rather than the tangible impact an analytical insight has on an investment decision.
• AI-Surfaced Attribution: Traditional email open rates and portal engagement metrics are declining as automated scrapers become the primary readers of research content. The sellside is adopting tracking mechanisms to verify which specific user queries surfaced their data, whether their analysis was cited in internal LLM answers, and when a human analyst clicked through to the underlying research.
• Technical Infrastructure Requirements: Buyside firms increasingly judge sellside partners on the quality of their Model Context Protocol (MCP) infrastructure and semantic indexing layers. Research providers that fail to deliver clean, AI-ready data streams face rapid exclusion from automated institutional workflows.

 

Regulatory Realignment: The MiFID II Reversal

The regulatory landscape has reached a clear turning point. Revised unbundling rules in the European Union are active, and the UK Financial Conduct Authority (FCA) has established its updated framework. The primary hurdle to returning to commission-bundled research models is now behavioural rather than legal.

Virtually all tier-one global asset managers are hesitant to move first due to concerns over client perception and negative media commentary regarding passed costs. However, smaller asset managers under severe P&L pressure have already transitioned, setting up a “trickle-up” trend that is expected to drive wider institutional adoption by an industry-wide target date of January 2027.

The operational flexibility of commission-funded models is highly compelling. The FCA’s framework permits mid-year budget adjustments to handle unexpected macroeconomic volatility. Under a strict corporate P&L model, securing additional unbudgeted research funding from an internal CFO mid-year is immensely difficult.

Furthermore, global managers face growing institutional pressure over inconsistent client treatment. Charging US clients for research via Section 28(e) while absorbing those same costs on P&L for European accounts creates an unsustainable operational double standard. When investment managers present detailed cost dashboards and rate cards to asset owners, the consistent response from pension trustees is that they care about net fund performance, not granular operational accounting. A minor charge of a few basis points against the fund for high-quality research is negligible if it drives outperformance. Additionally, new EU rules require regulated firms to formally assess the absolute quality and utility of their research input, regardless of whether it is funded via P&L or a Client Commission Sharing Agreement (CSA).

 

Market Outlook and Strategic Imperatives

As capital markets converge on the January 2027 regulatory target, institutional participants must execute on three strategic priorities:

1. Prioritise LLM ready content  Over Static PDFs: Sellside institutions must quickly package their analytical insights into semantically tagged, MCP-compliant data architectures. If written material cannot be programmatically parsed by buyside algorithms, it will be excluded from modern workflows.
2. Establish Clear AI Attribution Frameworks: Research providers must implement robust logging systems to capture when their intellectual property is ingested, queried, or cited within client LLMs. This data is essential for protecting IP and ensuring fair attribution during broker votes.
3. Prepare Client Communication for CSA Transitions: Asset managers should take advantage of current benign market conditions to transition client accounts toward flexible CSA models. Restructuring research funding mechanisms is far easier now than during a market downturn.

 

Aligning Infrastructure with Innovation

Adapting to this algorithmic and regulatory environment requires breaking down technology silos across the entire capital markets enterprise. Legacy, fragmented systems struggle to manage the seamless data flow required for AI-ready research authoring, granular client engagement tracking (CRM), algorithmic sales and trading, optimised corporate access, and advisory workflows in investment banking.

Firms can navigate this transition by adopting an unfragmented, unified end-to-end architecture that connects content creation directly to monetisation. When authoring tools feed compliance-ready metadata straight into client engagement systems, sales teams gain a clear overview of client interests, trading desks optimise targeted liquidity provision, corporate access teams align registries efficiently, and investment banking teams execute transactions with deep data-driven context.