AAAnow, 25 years of innovation

The use cases · Public companies

Shareholder confusion.

A public company’s published record now answers investor questions through AI, whether or not anyone asked it to.

Results announcements, legacy investor pages and guidance documents stay available long after they are superseded. AI systems read them all with equal confidence, and answer with whichever they find.

ReferenceUC/2026/01
Updated12 August 2026
SectorPublic companies
ExposureMisinformation, misrepresentation
Where it startsThe investor-facing estate

01Where it begins

A superseded figure, an old guidance range or a withdrawn statement remains published somewhere in the estate: an archived results page, a forgotten PDF, a regional site. AI systems treat it as part of the current record.

02How it compounds

Each repetition reinforces the position. Analysts, advisers and retail investors hear the same superseded answer, and the correction has to reverse an established position rather than set a correct one.

03What AI Readiness changes

AI Readiness addresses the cause rather than the symptom: the organisation’s own online position.

Discovery

The investor-facing landscape is discovered in full: each website, page and PDF, known or forgotten.

A governed position

1 current position stands. Superseded material is identified and brought under control.

Representation monitored

What AI systems say about the company is monitored against the governed record, continuously.

Next step

Talk it through, in your terms.

The briefing sets this exposure in your sector’s terms, and the discussion follows your questions.

An illustrative scenario. No client is referenced, and no organisation has been assessed.

Questions

The questions this scenario raises.

Asked by boards in this position, answered directly.

An AI tool told a shareholder something our filings do not say. Where did that come from?

Almost always from your own estate: an old announcement, a stale investor page, a forgotten microsite still live and still readable. AI systems weigh what is published under your name; they do not know which of it you still stand behind.

Is this our problem or the AI vendor's?

The reading is theirs; the source material is yours. Root cause analysis across misinformation events shows the organisation's own online position is overwhelmingly where the error begins, which also means it is where the fix is available.

What is the fastest way to know our exposure?

Have the estate read the way AI reads it. The landscape stage collects what exists under your name and grades it against the AI Readiness maturity model, so investor-facing contradictions surface as findings rather than phone calls.

Can the confusing material simply be taken down?

Where it no longer earns its place, yes: route removal is part of remediation. The order matters though: landscape first, so removals are chosen on evidence rather than on whichever page caused this week's question.