Who we serve
Public Figures
The achievable objective is proportion and context, not disappearance.
Public figures have the hardest version of this problem and the clearest one. Hardest because the legal protections are thinnest for public figures, which is why the work turns on proportion and on the record you build. Clearest because that constraint eliminates the options that waste other clients' money.
The work is therefore proportion. A public record is going to contain the difficult episodes; the question is whether it contains everything else at comparable weight. A great deal of what damages public figures is technically accurate and materially misleading: a quotation stripped of its question, a settlement described without its terms, a decade-old position presented as current. None of that is removable and all of it is addressable by making the fuller version equally findable. The AI dimension is acute here: models compress long careers toward the most distinctive episode rather than the most representative one.
See an exampleWhy this behaves differently
-
Weaker legal protection
Actual malice standards and reduced platform protections. The available levers are editorial and structural rather than legal.
-
Accurate but misleading is the norm
Stripped quotations, missing outcomes, outdated positions presented as current.
-
Litigation is a visibility decision
Filing creates a durable public record. That trade-off has to be quantified before it is made.
What usually applies
-
Earned Authority
Substantive published work that supplies the rest of the record at comparable weight to the difficult parts.
-
AI Answer Correction
Where model summaries of a long career are unrepresentative or omit resolutions and outcomes.
-
Search Result Displacement
Restoring proportion in position, where the material itself is permanent.
An example
The job is proportion.
A public figure has a real record: offices held, dates, decisions, coverage that was accurate when it was written. That material has slipped down. What has risen is a cluster of older attack pieces, reprints and short items that are easier to index than the full arc. AI answers compress a career into its loudest paragraph, which is a different thing from its most complete one.
A reader willing to go to page two can find the accurate reporting. A reader who stops at the first screen, or at a model’s summary, cannot. The brief is completeness: some of the hostile pieces are commentary on things that happened, and accurate coverage stays up.
What changes is order and completeness. Current, sourced accounts of the actual service and dates move into the same view as the older material, so an AI answer can begin with the office and the span of time and take the disputes in the order they occurred. The hostile pieces remain on the web. They stop being the entire available story.
Proportion is the clean result in this category.
An illustration of how this usually goes, not a client matter. All seven examples · Who we serve
Asked in these matters
Can you fix my Wikipedia entry?
What we can do is establish that the reliable sources it draws on are accurate, which is the route its own editors are meant to follow. Wikipedia has conflict-of-interest rules, so the work goes through its channels: sourced corrections, proposed openly.
Tell us where you are.
One conversation, in confidence, with a clear reading of what is there, including when the answer is that nothing needs doing.
In confidence. We reply within one business day. The preliminary review is free, and you decide whether to continue.