Straight answers about Beacon.

Beacon provides independent value assurance for important business cases and initiatives.

It continuously reviews the information available, challenges assumptions, identifies emerging risks and inconsistencies, and highlights what may require management attention. The aim is simple: help decision-makers see important issues early enough to act.

Beacon is designed to work across sectors and different types of decisions. That can include major capital projects, portfolio investments, new technology ventures, acquisitions, strategic programmes and organisational transformations.

The subject matter changes, but many of the questions are similar: Is the case still sound? What has changed? Which assumptions matter most? Where is the evidence weak? What could undermine the expected outcome?

No. Beacon does not replace project management, portfolio management or reporting systems.

Those systems help teams manage and report the work. Beacon provides an independent perspective on the business case behind it and asks whether the available evidence still supports the conclusions being presented to decision-makers.

No. Beacon supports human decision-making; it does not replace it.

It identifies issues, challenges assumptions and brings relevant evidence together so that executives and investment teams can make better-informed decisions. Accountability for the decision remains with your organisation.

Most copilots are designed to help a team do the work: finding information, summarising documents, drafting content or answering questions.

Beacon is designed to challenge the work. It looks across the available evidence, tests whether important conclusions remain supported, identifies contradictions and changing assumptions, and highlights issues that may deserve management attention.

Yes. Beacon uses modern AI models as part of its reasoning capability, but the model itself is not the product.

Beacon combines AI with a professional assurance methodology, structured evidence, continuity between reviews, safeguards and purpose-built software. This turns the underlying AI into a repeatable independent value-assurance capability. As AI models improve, Beacon can benefit from those improvements without changing the purpose of the service.

Yes. Organisations with the right expertise and resources could build their own value-assurance capability.

Doing so means developing and maintaining not just an AI application, but also the assurance methodology, evidence model, safeguards, security, interfaces and operating processes around it. Beacon provides that as a maintained capability, allowing clients to focus on the decisions being reviewed rather than building and operating the assurance system themselves.

Yes. Beacon's strengths compound as more reviews are completed.

Recurring signals, patterns and early warning indicators can be recognised across different types of business cases. These lessons help Beacon ask better questions, identify relevant issues earlier and distinguish meaningful signals from routine project noise.

Today, this learning is curated by the Beacon team and incorporated into the review methodology and system. Over time, more of it can be built directly into Beacon's agents. Confidential client information is not shared between organisations.

This is another difference from a general-purpose AI conversation: each Beacon review benefits from a growing, specialist body of review experience.

AI is making software easier and cheaper to create. We don't believe the software alone is where Beacon's value lies.

The value comes from the combination of methodology, technology, accumulated learning, independent challenge and professional judgement. Clients are not simply buying access to another AI application. They are gaining a continuous independent value-assurance capability designed to bring important signals to decision-makers early enough to matter.

Beacon can work with the material already used to manage and govern a business case, including investment papers, business cases, plans, schedules, financial information, risk registers, decision records, progress reports and other supporting material.

Where relevant, external information can also be considered when it could materially affect the case.

Usually not. Beacon is designed to sit alongside existing governance and decision-making processes rather than replace them. It works with information organisations already produce and adds an independent perspective.

No. Continuity is an important part of Beacon.

Beacon maintains its understanding of the business case as new information becomes available. This allows it to assess not only the current position, but also what has changed, which assumptions have moved, whether previous concerns have been addressed and whether new issues are emerging.

Yes. Beacon is designed around evidence rather than an open-ended AI conversation.

Findings are linked to the evidence behind them, allowing important conclusions to be traced back to their source material. When new evidence changes an earlier conclusion, Beacon preserves that history rather than simply overwriting it.

Beacon uses large language models, so hallucination risk cannot be eliminated completely. But the risk is materially reduced compared with using a general-purpose AI such as ChatGPT, Claude or Gemini on its own.

That is because Beacon does not simply ask an LLM for an answer. The model operates inside a governed professional review system. Material findings are grounded in project evidence, important conclusions can be traced back to their sources, and Beacon keeps evidence, interpretation and uncertainty explicit. Its review methodology and safeguards constrain how the AI is used rather than relying on an open-ended conversation.

No. Independent value assurance should not manufacture concerns simply to demonstrate value. If the available evidence supports the business case and no material issue has emerged, Beacon should say so.

Equally, when evidence is incomplete or contradictory, Beacon should make that uncertainty visible rather than pretending to know more than it does.

It is deliberately a combination of software and professional expertise.

Software provides continuity, structured analysis and the ability to review large amounts of information consistently. Professional assurance methodology and expertise provide the judgement and challenge around it. This makes Beacon closer to an ongoing value-assurance capability than either a traditional software product or a one-off consulting review.

No. Some decisions require specialist technical, legal, financial or other assurance. Beacon can help identify where deeper expert review may be warranted, but it is not intended to replace specialist professional advice.

Yes. Beacon is designed to work with confidential and business-critical information.

For particularly sensitive cases, Beacon also provides the option to anonymise information before it reaches the AI layer, allowing sensitive identities and details to remain protected while the underlying business case can still be reviewed.

Each client organisation has its own separate workspace.

Access controls ensure users can only access information belonging to organisations they have been authorised to join. Documents are stored privately and reviews are performed separately for each business case. Client information is not pooled or shared between organisations.

Beacon's database and document storage are hosted on managed cloud infrastructure in the EU.

Some information is processed by external AI services when Beacon performs its analysis. For particularly sensitive cases, Beacon's anonymisation option can remove identifying information before it reaches the AI layer.

Access to a Beacon workspace is controlled by organisation membership. Users sign in securely and can only access organisations they have been authorised to join.

Where Beacon personnel require access to support a client engagement, access is provided to named individuals within that client's workspace.

Where Google Drive integration is enabled, you share a selected folder or Shared Drive with Beacon's own Google identity on a read-only basis.

Beacon does not need access to an individual's personal Google account and can only read information that has explicitly been shared with it.

Yes. A business case can be deleted from Beacon, removing its documents, evidence, findings and review history from the active Beacon environment.

As with most managed cloud services, deleted information may remain temporarily in routine infrastructure backups before those backups expire.

Beacon does not require a perfect or complete data set.

Existing business-case and governance material is usually enough to establish an initial view. Missing or weak evidence can itself be important information, and Beacon can identify where additional evidence would materially improve the review.

No. What matters more is the importance and complexity of the decisions being reviewed.

Beacon becomes particularly valuable when information is spread across multiple sources, circumstances change over time, management wants an independent perspective, or an organisation has multiple investments or initiatives that would benefit from a consistent review approach.

Yes. Starting with a single live business case is often the simplest way to evaluate Beacon. It allows you to see how Beacon works with your own information and compare the resulting review with your existing governance and assurance processes.