THE MARTHA INQUIRY™
Media
Accountability &
Real-time
Truth-seeking through
Honest
Algorithmic review
Going to the Source in the Algorithmic Era
The permanent legal architecture for media integrity and public discourse.
MARTHA GELLHORN'S LEGACY
Honoring Martha Gellhorn:
From Bypassing Censorship to Breaching the Algorithmic Black Box
Martha Gellhorn was born in St. Louis, Missouri, in 1908. She was a war correspondent for nearly sixty years, covering conflicts from the Spanish Civil War to the American invasion of Panama. She reported from the front lines, not from the briefing room.
She waged a lifelong campaign against what she called "managed news," the practice of allowing power to control the narrative. She believed that truth required going to the source, regardless of the obstacles. She went undercover during the Great Depression to report on the lives of the unemployed. She stowed away on a D-Day hospital ship when women were barred from the front lines. She landed at Normandy on June 6, 1944, and was the only female journalist to do so.
She was among the first journalists to document the liberation of Dachau. She wrote plainly about what she saw, without euphemism, without the managed language of official communiques.
Her legacy demonstrates that truth requires going where the story lives, regardless of the obstacles. The new managed news operates through algorithms that silently amplify, suppress, and distort public discourse. The MARTHA Inquiry continues her legacy by going directly to the source of these algorithmic systems.
THE ALGORITHMIC THREAT
A misinformation classifier does not arrive in a newsroom as a suppression order. It appears as a confidence score. The approved outlet reaches the feed. The unfamiliar source does not. The ranking becomes the reach. The reach becomes the silence.
The same is true for content moderation systems, trust-and-safety classifiers, spam filters, and distribution algorithms. When the logic is buried in a black box, the harm is invisible. The journalist has no explanation. The regulator has no violation to cite. The platform has no liability on the books.
But the liability is real, and the harm is now documented.
This is the pattern Martha Gellhorn spent her career fighting. Managed news does not require a blatant lie. It only requires a filter. The filter decides what gets seen, what gets boosted, and what quietly disappears. It does not need to be malicious. It only needs to be confident.
The algorithm is our new filter. When a misinformation classifier flags accurate reporting as unreliable, it is not committing a factual error. It is operating within the same institutional bias that Gellhorn documented from the front lines. Trained on data that treated outsider voices as risks, the model inherits the conclusion. It applies it at scale, in milliseconds, without a byline.
Classifiers trained on raw engagement systematically penalize stories from minority-owned outlets, community publications, and independent journalists. Moderation systems trained on historical takedowns replicate those exact enforcement patterns against the same communities. News ranking algorithms promote the dominant and suppress the marginal.
The harm is structural, not isolated. Media models that share training data, architecture, or third-party dependencies with foundation models propagate bias across the entire digital ecosystem. A single vendor's classifier becomes every platform's suppression.
That suppression is now an instrument of state action. Investigations by courts and state attorneys general have exposed how platform moderation, when coordinated with government agencies, transforms private curation into a First Amendment violation. The question is no longer whether the algorithm suppressed the story. It is whether the state asked it to.
The journalist has no way to see the score. The regulator has no way to audit the weights. The platform has no way to prove the model was governed.
Until now.
THE MARTHA INQUIRY SERVICES
WHAT WENT IN
1. Data Provenance Audit
We document the provenance of the training data behind every misinformation classifier, content moderation model, and news distribution algorithm. We verify every data point at the source, trace what the model encoded at the weight level, and produce a baseline record that satisfies the evidentiary standards of First Amendment claims, defamation actions, and state anti-SLAPP review.
Powered by COPERNICUS Canon™.
2. Synaptic Silencing Audit
Before you can govern a media model, you must identify what needs to be silenced. We document what the model carries at the weight level before it reaches the public. We cross-reference the developer's AB 2013 disclosure, record the fine-tuning history, and produce the pre-deployment baseline record.
Powered by DigitalRAS™, Module 1.
WHAT IT DOES
3. Misinformation Classifier Audit
We audit the classifiers that flag, downrank, or demonetize content as misinformation. We verify that the classifier does not systematically penalize accurate reporting from minority-owned outlets, community publications, or independent journalists. We document the pattern when the model dismisses accurate claims as false.
Powered by DigitalRAS™.
4. Content Moderation Algorithm Audit
We audit the moderation systems that determine what remains published and what is removed. We verify that enforcement is consistent across outlets, that it does not disproportionately target protected speech, and that the appeal pathway is operational. When a moderation system is found to be discriminatory, we document the finding in a litigation-ready format.
Powered by DigitalRAS™.
5. News Ranking and Distribution Audit
We audit the ranking and recommendation algorithms that determine what the public sees. We verify that the model does not default to the outlets that were already dominant, and that the training data does not contain proxies for historical platform favoritism. We document the pattern when the model systematically suppresses the same sources over time.
Powered by DigitalRAS™.
6. Journalist Surveillance Audit
We audit the tools that monitor journalists. The tracking of sources. The monitoring of communications. The use of facial recognition at protests. The geofence warrants that sweep up reporters. When a journalist is surveilled by a system they were never told about, we document the violation.
Powered by DigitalRAS™.
7. Synthetic Media and Deepfake Audit
We audit the models that generate synthetic media of journalists. Deepfaked footage. Cloned voices. Fabricated quotes. We trace the model, attribute the output, and document the harm to the journalist and to the public record.
Powered by DigitalRAS™.
8. Inference Safeguard Validation
We validate that human review requirements for final takedown, demonetization, and account suspension decisions are actually operational. The system should hold irreversible actions for human review while allowing routine distribution to proceed. We document whether the verification layer exists, whether it functions, and whether it was bypassed.
Powered by DigitalRAS™.
HOW FAR IT REACHES
9. Systemic Correlation Assessment
We measure whether the media model shares training data, architecture, or third-party dependencies with widely deployed foundation models. We track behavioral drift across news cycles and generate the substantial-modification evaluation.
Powered by DigitalRAS™.
10. Vendor Disclosure and Accountability
We map the algorithmic supply chain behind the platform's moderation and distribution tools. We demand vendor disclosure of model logic, training data provenance, and fairness audits. Where vendors refuse disclosure, we document the refusal as evidence of concealment.
Powered by Regulatory Judo™.
WHAT WE DO ABOUT IT
11. Proactive Compliance Architecture
For platforms and publishers that want to govern the system before the state Attorney General or the FCC does.
We establish a Media Algorithmic Integrity Protocol that governs the design, deployment, and monitoring of moderation, classification, and distribution algorithms. This includes pre-deployment bias testing against protected speech, ongoing model drift monitoring, journalist grievance pathways that bypass the algorithm, mandatory human-in-the-loop requirements for all takedown decisions, and First Amendment, Section 230, and state readiness.
Powered by DigitalRAS™.
12. The Martha Redress Initiative
For journalists, outlets, and advocacy organizations.
We provide a Probabilistic Harm Audit that uses econometric counterfactual baselines to isolate the Algorithmic Increment of Harm in a specific suppression, takedown, or demonetization. This quantifies the harm for First Amendment claims, defamation actions, anti-SLAPP review, and impact litigation.
Powered by The Right to Be Probable™.
THE FRUIT OF THE POISONOUS MERKLE TREE™
The remedy doctrine. What happens when the story is suppressed.
A misinformation classifier does not produce a single suppressed story. It produces a tree of downstream consequences. The flag leads to the downrank. The downrank leads to the lost reach. The lost reach leads to the lost revenue. The lost revenue leads to the layoffs. The layoffs lead to the closed outlet. The closed outlet leads to the community that no longer has a local paper. The silence follows the community for a generation.
Every branch of that tree is tainted by the original algorithmic error. And every branch is still bearing fruit.
The Fruit of the Poisonous Merkle Tree™ treats algorithmic taint as a traceable structural condition. The doctrine provides a five-part remedy that goes beyond monetary damages.
Algorithmic Disgorgement. The platform must surrender the outputs it derived from the tainted model. Not just the classifier. Every flag, every downrank, every demonetization, every suspension that flowed from it. The taint travels. The disgorgement follows the taint.
Forced Structural Retraining. The platform must rebuild the system that produced the harm. Not patch it. Not audit it. Rebuild it. The training corpus, the moderation thresholds, the appeal pathway, the human review workflow. The correction is not a memo. It is a new architecture.
Downstream Notification. Every journalist, outlet, and publisher whose work was suppressed by the tainted system must be notified. The outlet that was demonetized knows. The freelancer whose story was downranked does not. The community newspaper that lost its distribution does not. The doctrine requires the platform to find them and tell them.
Supply-Chain Traceability. Every downstream distributor, aggregator, and licensee that received the tainted ranking must be identified. The search engine. The social feed. The news aggregator. The ad exchange. The taint travels. The remedy maps where it went.
Harm Apportionment. The harm is not the platform's alone. The model developer built the classifier. The integrator deployed it. The platform operated it. The advertiser funded it. Every party in the chain is assigned a share of the remedy proportionate to the role it played. This is where the Right to Be Probable™ does its work. The Algorithmic Increment of Harm isolates what each party contributed.
The doctrine is not punitive. It is structural. The taint travels. The remedy follows.
THE MARTHA LEGAL FRAMEWORK
The First Amendment. The constitutional protection against government abridgment of speech and the press. The anchor for every claim that a platform's moderation decisions, when coordinated with the state, constitute state action.
Section 230 of the Communications Decency Act. The federal statute that shields platforms from liability for user-generated content while preserving their right to moderate. The statute that defines the boundary between platform immunity and platform responsibility.
State Anti-SLAPP Statutes. The state laws that protect defendants from meritless lawsuits designed to chill speech. The procedural shield for journalists and publishers.
State Shield Laws. The state statutes that protect journalists from being compelled to reveal their sources. California, New York, and most states have shield laws. The federal PRESS Act would create a national shield.
Defamation Law. The state and federal doctrine that governs false statements of fact that harm reputation. The anchor for claims against synthetic media that fabricates quotes or imputes misconduct.
The EU Digital Services Act. The European regulatory framework that requires transparency in content moderation and provides appeal rights for users and publishers.
The Fruit of the Poisonous Merkle Tree™. The remedy doctrine. Five-part structural remedy for algorithmic taint in media distribution.
Regulatory Judo™. The disclosure engine. The right to know whether AI was involved in the suppression.
Eight instruments. One record.
WHO THIS SERVES
The MARTHA Inquiry serves the entire media ecosystem. Platforms and publishers that want to build defensible moderation systems. Journalists and outlets that want to hold those systems accountable. And the attorneys, regulators, and advocates who bring the cases to a legal forum.
For Platforms, Publishers, and Broadcasters
Platform Policy Teams. Trust and Safety Leaders. Newsroom Counsel. Broadcast Standards Teams. Ad Exchange Compliance Teams.
If you want to audit your moderation, classification, and distribution systems before the state Attorney General or the FCC finds them, we apply the DigitalRAS forensic engine, the Misinformation Classifier Audit, and the Content Moderation Algorithm Audit to identify and remediate exposure before it becomes a liability.
For Media Tech and Moderation Vendors
Content Moderation Vendors. Misinformation Detection Developers. News Ranking Platform Builders. Product Counsel. Regulatory Affairs Teams.
If you are building the classifiers that platforms deploy, you carry obligations under the First Amendment, Section 230, and the EU Digital Services Act before your model ever reaches a newsroom. We built the weight-level governance architecture that makes your model defensible at the point of sale and at the point of subpoena.
For Journalists, Outlets, and Press Freedom Organizations
Journalists. Editors. Freelance Reporters. Independent Publishers. Community Newspapers. Press Freedom Organizations. Media Unions. State and Federal Regulators.
If your story was flagged, downranked, demonetized, or suppressed by a system you were never allowed to see, you may have a claim. The record already exists. Let us read it.
For Media Law Attorneys & Plaintiffs' Firms
We serve as co-counsel and forensic support on First Amendment, defamation, and anti-SLAPP cases that turn on what the algorithm did and how the record proves it.
THE MARTHA ECOSYSTEM
The MARTHA Inquiry is the media and journalism configuration of DigitalRAS. It is one sentry in a coordinated system of algorithmic accountability. Two related engines extend its reach across the cultural record.
THE HATTIE TAKE™
The Creative & Entertainment Sentry
Where MARTHA audits the algorithm that determines what the public reads, watches, and believes, HATTIE audits the algorithm that replicates the performer. Both govern the systems that shape the cultural record. HATTIE protects the person who made the work. MARTHA protects the truth the work was supposed to tell.
REGULATORY JUDO™
The Disclosure Request Engine
LAUNCHING SOON
The journalist has the right to know whether AI was used to flag, downrank, or suppress the story. Regulatory Judo generates the legally compliant disclosure request and builds the record that feeds the pattern that feeds the case.
→ Join the Regulatory Judo Waitlist
THE CULTURAL RECORD LIFECYCLE
Perform. Publish. Attribute. Disclose.
Two sentries. One lifecycle. One record.
WHY THIS IS 1 OF 1
Most governance programs treat AI risk as a policy problem. They produce frameworks, checklists, and statements of principle. The policies exist. The training exists. The safeguards exist. None of it is enough.
The Architecture treats AI risk as an architecture problem. It governs the model, the process, and the record simultaneously. It accounts for both the technical failure and the human oversight that permits it.
System failures require a multi-layered explanation. The technical breakdown pairs with automation bias. It compounds with siloed review. It feeds on a lack of imagination regarding system vulnerabilities. It thrives under an incentive structure that prioritizes speed over verification.
The model fails because it executes inside an operational framework that was never built to catch it.
The Architecture is built at the weight level. It produces a record that is forensically defensible. It is designed to be tested.
No other legal framework currently combines misinformation classifier forensic auditing with First Amendment enforcement, content moderation review with anti-SLAPP protection, journalist surveillance auditing with shield law defense, and proactive platform compliance with journalist-side and press freedom litigation support.
No other legal framework combines weight-level media forensics with a dedicated media accountability practice. MARTHA produces the record. Narrative Strategy & Impact Litigation enforces it.
The MARTHA Inquiry is not a generic AI ethics framework. It is a weaponized legal instrument built specifically for the media sector, where algorithmic suppression is already widespread, largely unregulated, and generating First Amendment exposure at an accelerating rate.
It is the only sentry in the portfolio named for a war correspondent. Martha Gellhorn spent sixty years going to the source, past the briefing, past the managed narrative, past the gatekeepers who said she could not be there. The MARTHA Inquiry does the same work at the weight level. It goes to the model. It goes to the training data. It goes to the inference log. It does not accept the platform's summary. It goes where the story lives.
Thirteen sentries. One engine. One record.
The press is already scaling. The MARTHA Inquiry ensures you are not left behind.
The difference does not show up in a slide deck. It shows up in discovery.
