What makes board decision-making effective in 2026?

 The conditions under which boards make decisions have shifted materially in the past three years.

Agenda volumes are rising: the Spencer Stuart UK Board Index reports that the average FTSE 350 board now handles more agenda items per meeting than at any point in the previous decade, driven by regulatory complexity, geopolitical exposure, and the pace of technology change. At the same time, the personal and institutional consequences of getting decisions wrong have grown. The UK Corporate Governance Code (2024 revision) tightens expectations on decision documentation, and the EU AI Act is introducing a new layer of compliance obligations for any board that relies on AI-assisted analysis in its decision process. 

Against that backdrop, 71% of dealmakers surveyed by FT Longitude in 2026 believe that firms that ignore AI today will not be able to compete within five years.[2] That figure was calibrated to transaction teams, but the underlying logic applies with equal force to the board: the standard of analysis, preparation, and decision capture that was adequate five years ago is no longer the ceiling. The boards that will navigate this period well are those that have treated decision quality as a discipline worth improving, not a function of who happens to be in the room. 

Effective board decision-making in 2026 is the disciplined combination of high-quality information, structured deliberation, clear accountability, and a well-judged use of AI as decision support rather than decision substitute. 

This piece works through four observable disciplines that distinguish effective boards, four failure modes that recur across organisations regardless of their governance maturity, and a meeting design that ties them together. The aim is practical: a Chair or Company Secretary should be able to identify two or three habits worth changing before the next meeting. 

What effective board decision-making looks like 

Effective board decisions are not the product of exceptional individuals or unusually good chemistry around the table. They are the product of four observable disciplines, each of which can be built into the structure of a meeting and the design of the preparation that precedes it. 

Information quality and access 

Information quality in a board context is the joint property of being accurate, complete, timely, and decision-relevant. A paper that is accurate but not decision-relevant wastes the board's time. A paper that is decision-relevant but arrives 48 hours before the meeting produces a conversation, not a decision. The PwC Annual Corporate Directors Survey (2024) found that a meaningful share of directors report feeling under-informed at the point of decision, with the most cited cause being papers that are too long, too late, or too disconnected from the specific question the board is there to answer.

The practical fix is not shorter papers but better structured ones. Each paper in the decision pack should carry a one-page executive summary that answers three questions: what is the board being asked to decide, what are the key assumptions, and what is the range of reasonable outcomes. The full paper follows for those who need it. Pre-circulation should happen at least seven calendar days in advance.

Information quality is also information control. Pre-circulating a paper that contains revised capital allocation scenarios or a confidential acquisition rationale requires role-based access permissions, watermarking, and data residency that sits outside the reach of foreign surveillance legislation.

Structured deliberation 

Structured deliberation is the deliberate sequencing of how the board moves from information to decision: a clearly framed opening question, a period of individual reflection, non-executive director (NED) responses before executives, and a verified conclusion before the decision is recorded. The research on why this sequencing matters is not new. Cass Sunstein, Olivier Sibony, and Daniel Kahneman's work on noise, published in 2021, demonstrated that unstructured group deliberation systematically amplifies the first confident voice in the room, a finding that translates directly to the dynamics of a board meeting where the CEO or CFO typically speaks first after a management presentation.

The practical mechanics are straightforward and do not require a cultural transformation. Frame the agenda item as a decision question rather than a topic heading: 'Should we proceed with the proposed Iberia expansion at the Q3 capital threshold?' is a different conversation from 'Iberia expansion update'. Ask non-executives to respond before executives, reducing the anchoring effect of the management team's opening position. Before the decision is recorded, the Chair or Company Secretary paraphrases the conclusion and confirms shared understanding explicitly: 'We are recording approval of the Iberia expansion at the Q3 capital threshold, with a review at the January meeting. Is that the shared understanding?' The Russell Reynolds Global Leadership Monitor has identified precisely this kind of deliberation discipline as one of the strongest correlates of high board effectiveness ratings across its global dataset.

Clear accountability for outcomes 

Accountability in the context of board decisions means three specific things: who owns the outcome, by when, and with what reporting cadence back to the board. The UK Corporate Governance Code (2024) is explicit on decision documentation, requiring that boards maintain a clear record of the basis on which significant decisions were taken. The practical failure that EY's Center for Board Matters has documented repeatedly is not that boards fail to decide, but that the decision record does not survive the meeting with sufficient precision to hold anyone accountable later.

The fix is mechanical. Every decision should leave the meeting with a named owner, a deadline, and a date on the next agenda when progress will be reviewed.

The role of AI in board decision-making 

The 2026 FT Longitude survey of 1,000 dealmakers, commissioned by Datasite, asked respondents to identify the top attributes they require from AI when it supports decision-making. The results were: accuracy (71%), security (70%), reliability (58%), speed (57%), and compliance (44%).The ordering is significant. Security ranks second, not fourth. For boards handling confidential strategic material, the criterion that distinguishes board-grade AI from general-purpose AI is not primarily its analytical capability; it is whether it can be trusted with the material at all. The same expectation is forming around board-level decision-making more broadly.

Three concrete uses of AI in board decision-making have emerged as practically valuable and governance-compatible. Document Copilot lets a director or Company Secretary ask focused questions of a 200-page paper pack and surface the decision-relevant content in minutes rather than hours. The AI does not produce a recommendation; it surfaces the specific section of the valuation report, the risk register entry, or the regulatory opinion that is relevant to the question the board is trying to answer. Summarising prior decisions and meeting-to-meeting context maintains continuity across a board's work without relying on any individual's memory. And AI-assisted agenda and documentation drafting frees preparation time for analysis rather than administration. 

The clearest evidence that the decision itself remains the board's comes from the same survey: only 22% of dealmakers said they would follow an AI-generated recommendation on whether to sign a deal. That figure was consistent across AI-enthusiast and AI-cautious respondents. The pattern is not reluctance; it is calibration. Even practitioners who use AI extensively for analysis do not delegate the judgement call to it. That is the right frame for boards: AI as the mechanism that improves the quality of the information on which the board exercises its judgement, not the substitute for that judgement. 

The EU AI Act introduces a relevant regulatory frame for boards that are thinking about how to govern their AI use. Systems that support decisions with significant consequences for individuals or organisations may be classified as high-risk under Articles 6 and 7, which triggers documentation, transparency, and human-oversight requirements. Those requirements align naturally with good governance practice: a board that already maintains an immutable decision record and requires human sign-off on AI-assisted analysis is well-placed relative to those that have adopted AI tools without considering their governance implications. 

The 62% of dealmakers who said human-only decision-making is no longer defensible in complex deals is a data point about M&A practice. Its relevance to board work is directional, not literal: the principle that AI-assisted analysis is becoming the standard against which unassisted analysis will eventually be measured is one boards should register, even if the pace of adoption at board level lags the pace in deal teams. 

AI as the new pre-read: Document Copilot lets a board member ask focused questions of a 200-page paper pack and surface the decision-relevant content in minutes. It does not make the decision. It frees the time to make it well. 

Common decision-making failures boards should avoid 

The four failure modes below are not hypothetical. They appear in the governance research literature and in the disclosures and regulatory findings that follow high-profile board failures. Each is structural in origin, which means each is addressable by changing the mechanics of how the board prepares and deliberates, rather than by recruiting different people. 

• Ratification masquerading as deliberation. The board approves a decision the executive team had already taken. This is the failure that investor relations teams and proxy advisers track most closely, and it is the failure the OECD Principles of Corporate Governance are most explicitly designed to prevent. It tends to happen when the paper pack presents a single preferred option with supporting analysis rather than a range of options with competing analysis. The structural fix is simple: require the paper to present at least two realistic options and the board to record why the alternative was not selected. 
• Anchoring on the loudest voice. A single board member's view sets the direction of discussion before others have spoken. When that voice belongs to the CEO or CFO presenting the item, it is not malice; it is the natural dynamics of a room in which one person has more context than the others. Sunstein and Sibony's research on informational cascades in group settings documents this pattern in detail. The fix is sequencing: NEDs respond first, executives respond to them. 
• Ambiguous decision capture. The minutes record a discussion but not a clear decision, an owner, or a deadline. The next meeting then relitigates what was supposedly settled, which compounds over time into a board culture in which nothing is ever quite decided. ICSA's guidance on board effectiveness identifies ambiguous minute-taking as one of the most common contributors to governance dysfunction in otherwise well-structured boards.
• Structural under-use of AI for board preparation. The 2026 FT Longitude / Datasite survey found that 27% of dealmakers are still not using AI for board reporting and governance. At a point when the analytical tools available to board-level decision support have improved substantially, continuing to interrogate 200-page paper packs manually represents a structural choice with diminishing returns. The objection that AI cannot be trusted with confidential board material is a question of platform, not a question of AI in the abstract: it requires a board-grade platform with the right security credentials, not a decision to forgo the analytical support entirely. 

How meetings support better board decisions 

The meeting design choices that support good decision-making are not complicated. They require consistency of application more than sophistication of design. 

Meeting cadence should provide enough sessions to deliberate substantively on complex topics, while leaving sufficient time between meetings for individual preparation to be worthwhile. A board that meets too frequently trains its members to skim the papers; one that meets too infrequently accumulates business until each session is too crowded for deliberation. Most effective FTSE 350 boards settle in the range of six to eight scheduled meetings per year, with the ability to convene additional sessions for specific gating decisions. 

Every agenda item should be framed as a decision question before the meeting begins. This applies to both strategic items and what appear to be routine governance matters: a decision question forces the paper author to be explicit about what the board is being asked to approve, and it gives the Chair a clear basis for concluding the item.

During the meeting, the discipline of NEDs responding before executives is supported by Presenter Mode, which keeps confidential speaker notes and annexes off the shared screen and allows the Chair to manage the room without the room seeing the management position before the NEDs have formed their own. Before any decision is recorded, the Chair or Company Secretary verifies shared understanding by paraphrasing the conclusion explicitly. 

After the meeting, a short feedback exchange on the quality of the decision-making process builds a data set on whether the board is improving over time. The UK Corporate Governance Code's provisions on annual board evaluation provide the regulatory anchor for this feedback loop; ICSA's guidance on board effectiveness suggests that boards which track decision quality as a specific metric within their evaluation process report more consistent improvement than those that treat evaluation as a general exercise. 

How Sherpany improves board decision-making 

Sherpany is a Swiss-headquartered meeting management platform used by more than 450 boards and 20,000 directors and executives each week, with a 98% annual renewal rate. The customer base is concentrated in regulated industries: banking, insurance, financial services, pharma, manufacturing, energy, aviation, and transport and logistics. 

• Prepare: Topic Hub centralises proposed decisions ahead of the meeting, giving functional leads and board members a structured channel to submit agenda items in advance rather than introducing them at the last minute. Document Copilot lets directors interrogate long paper packs with focused questions, surfacing the decision-relevant sections without replacing the need to read the underlying analysis. The Meeting Preparedness Indicator signals to the Chair, before the session begins, which papers have been read and which board members may need additional time. 
• Meet: Agenda Builder structures every item as a decision question, with the stated purpose and expected outcome visible to all participants before the meeting opens. Presenter Mode keeps speaker notes and confidential annexes off the shared screen, maintaining the integrity of the in-room deliberation without requiring the Chair to manage it manually. 
• Follow through: Tasks and Decisions captures every decision against a named owner and a deadline at the point it is made, carries it forward automatically to the next agenda, and surfaces overdue items before the board reconvenes. The audit trail this generates records what was decided, by whom, and on what basis, which is the mechanism that makes the decision defensible to regulators, shareholders, or, in the worst case, a court. 

The platform's security and compliance infrastructure includes ISO 27001 certification (since 2018), SOC 2 Type II (since 2021), FINMA 2018/3 alignment, EU DORA readiness, and GDPR compliance, operating on AES-256 encryption at rest and TLS 1.3 in transit from Swiss data centres outside the scope of the US CLOUD Act. 

From agenda item to decision record: Every decision in Tasks and Decisions carries an owner, a deadline, and a follow-up date. The next agenda picks up where the last one left off, and the board stops relitigating what was supposedly settled. 

Elevate how your board approaches strategic decision-making today

The four disciplines described in this piece (information quality, structured deliberation, clear accountability, and a well-judged use of AI as decision support) are not theoretical. They are observable in the boards that consistently produce defensible, well-documented decisions and in the evidence base that tracks board effectiveness over time. None of them requires a change of personnel. Each requires a Chair and Company Secretary willing to treat the mechanics of decision-making as seriously as the substance of the decisions themselves. 

The forward-looking case for investing in these disciplines now is also clear. By 2030, the single biggest expected benefit of AI at the board reporting and governance stage is higher-quality decision-making, cited by 23% of the 1,000 dealmakers surveyed in the 2026 FT Longitude / Datasite research.[2] The boards that are already building the structural disciplines described here will compound those gains as AI-assisted decision support becomes the norm rather than the exception. 

One action worth taking this quarter: review the agenda from the last three meetings and count how many items were framed as decision questions rather than topic updates. The ratio is a reliable indicator of whether the board is deciding or just meeting. 

For a fuller treatment of the mid-year governance cycle, the companion guide The board's guide to strategic mid-year decision-making covers strategic evaluation, board composition, and M&A oversight in a single practical resource. Our recent strategic planning article addresses the structural role of the board in building the annual strategy process.

If you're ready to elevate the way your board makes strategic decisions, book a free consultation today and find out how Sherpany can help.