Board Software Vendors Race to Add AI Features as Governance Adopts a New Set of Promises

The board portal category has spent the first half of 2026 attaching the same set of artificial intelligence features to its products. AI-generated meeting minutes. Automatic transcription of board sessions. AI-drafted board reports. Engagement analytics interpreted by language models. The category roundups published this year by trade publications including Croclub, People Managing People, and Fellow.ai describe a vendor landscape in which Board Intelligence, Azeus Convene, OnBoard, BoardEffect, Govenda, and Diligent are competing on a feature set that, eighteen months ago, none of them was selling.

What the features actually deliver to a board, and where corporate governance still demands the kind of human judgment that the new tools do not provide, is the question the category has not yet answered with the same enthusiasm.

What is being marketed

The AI feature set in board software in 2026, as documented across the published reviews and vendor pages of the spring, is recognizably consistent across providers.

Board Intelligence, by the description in its current product positioning, markets AI-driven minute writing that generates meeting summaries automatically and an "Insight Driver" tool that flags where board reports may be missing key context. Azeus Convene, in the same period's reviews, markets automated video transcription and post-meeting recording that the vendor positions as a way to allow board members to revisit material after sessions conclude. OnBoard markets engagement analytics that, by the published descriptions, aggregate director activity into anonymized real-time insights for board secretaries. BoardEffect, Govenda, and Diligent have each layered comparable feature sets onto their existing governance platforms during the same period.

The vendors are not, by and large, claiming that these tools replace board work. The marketing language is generally careful, positioning AI features as administrative compression rather than substantive contribution. What is being sold is the recovery of hours from minute-taking, transcription, and report drafting, with the implicit promise that those hours can be redirected to the substantive work of governance.

The promise is, on its face, reasonable. The question is whether the implementation matches the framing.

Where the AI features genuinely help

The administrative load on company secretaries, governance professionals, and the staff supporting boards has been documented in trade press and professional surveys for many years. Minute-taking, in particular, has been a category of work that absorbs significant time and produces a document with substantial legal weight. The drafting effort has, traditionally, been an experienced human exercise. AI tools that produce a competent first draft of meeting minutes, leaving the experienced governance professional to review and refine rather than start from a blank page, are responding to a real problem and, in the cases where the implementations are competent, can compress the time required without compromising the integrity of the resulting record.

Transcription of board sessions, similarly, has a genuine case in environments where directors join remotely and may need to revisit specific portions of a discussion. The administrative value is concrete and the implementation risk is contained, provided the transcripts are handled with the same access controls as other board materials.

Engagement analytics, the third common feature category, have a more mixed case. The descriptions provided by vendors during 2026 emphasize aggregated and anonymized presentation of director activity, which is the framing that responsible governance practice would expect. The implementation question is whether the analytics, in practice, lead governance professionals to substantive improvements in board operations or to surface-level optimizations that focus on engagement metrics rather than on the underlying questions the board is meeting to address.

Where the limits sit

The board portal category's AI feature set is, in its current generation, an administrative compression tool. It is not, on the evidence available, a substitute for the elements of governance that depend on judgment.

A board exists to provide oversight, to set strategy, to hold management accountable, and to make decisions that the organization's stakeholders have entrusted to it. The work of doing these well includes reading materials carefully, asking questions that go beyond the surface of the papers, weighing trade-offs that the documents may not have framed explicitly, and committing to decisions whose consequences the directors must own personally.

None of that work is being performed by the AI features now being marketed into the category. A model that drafts a competent set of minutes is not making the underlying decisions. A model that transcribes a board session is not weighing the strategic trade-offs being discussed. A model that produces a board report from underlying data is not exercising the judgment that decides which data should have been collected in the first place.

This is not a criticism of the tools. It is a description of where they sit in the governance workflow. The risk, in our reading, is not that the AI tools are oversold by their vendors, who have been generally careful, but that the boards adopting them will treat the resulting time savings as evidence that less human attention is needed, rather than as an opportunity to apply the recovered attention to harder governance work.

The data protection question that has now arrived

A separate consideration, sharpened by the same week's news from the U.K. Information Commissioner's Office, is that AI features in board software now sit within the scope of the data protection obligations the ICO has clarified.

The ICO's May guidance, in the coverage published by Cybersecurity Insiders and other trade publications, identifies seven categories of AI-driven attack and maps them onto the existing security obligations of Article 32 of the U.K. GDPR. Board materials are, by their nature, among the most sensitive personal and corporate data an organization holds. AI features that process those materials are, by the ICO's framing, in scope of the same security expectations that apply to any other technology touching such data.

The implications for boards adopting AI-enabled portal features are not exotic. They are the same implications that apply to any AI deployment processing sensitive material: documentation of how the AI features handle the data, vendor representations about model training and security, confidence that monitoring extends to model inputs and outputs, and contractual provisions that allow the customer to demonstrate appropriate measures to a supervisor if asked. The vendors that have prepared for these questions in advance will, in our reading, fare better in procurement conversations during the second half of 2026 than vendors that have not.

What boards should ask

For boards and the governance professionals supporting them, the practical questions raised by the spring's news are concrete.

The first is what specific administrative work the AI features in their portal vendor actually compress, evaluated against the time spent on that work today. The honest evaluation is rarely identical to the marketing claim, and the trial of a feature on a meeting cycle or two reveals more than any vendor demonstration.

The second is how the AI features handle the underlying board materials, and whether the handling is consistent with the data protection obligations the board itself owes. The procurement conversation is the right place to raise this, in plain language, with documentation requested.

The third is what the board does with the time the features recover. The answer is not, in our reading, doing less governance. It is doing the harder parts of governance that the previous administrative load had crowded out.

The wider read

The board portal category in 2026 is racing to add features the boards adopting them will, in many cases, find genuinely useful, particularly for the administrative work that has long absorbed disproportionate time. The marketing is mostly careful and the technology is mostly competent. What separates the boards that benefit from the boards that do not will be the same thing that has separated good governance from less good governance for decades: whether the recovered attention is reinvested into the work that requires human judgment, or whether the time savings are simply absorbed elsewhere.

The features are arriving. The question is what the boards do with them.

Related reading

• How Board Software Supports Remote and Hybrid Meetings
• How Data Rooms Support Secure Information Handling Across a Business

Sources and further reading

• Diligent guide to AI board-meeting minutes tools
• Azeus Convene AI
• U.K. ICO guidance on AI-powered cyber threats

Impulsblog analysis is based on the published sources listed above and is current as of May 25, 2026.