Between the information and the decision.

Defacta is an independent verification layer between any text and the decision it informs, whether that's an article, a report, a filing, a social post, or an LLM output, AI-generated or written by a person.

It assesses the text and gives you a clear analysis of which claims are supported, how credible the sources are, and where bias, manipulation, or misleading framing may be shaping the message, so you can decide with the full picture in front of you.

How it works

Defacta works the way a careful analyst or fact-checker does, as a structured, multi-stage pipeline. LLMs speed up specific steps, but every step is anchored in fresh data retrieval and source-grounded analysis.

1. Claim verification. Every factual claim in the text is checked against current, credible sources.
2. Source credibility. Each source is assessed for its reliability and track record.
3. Manipulation and framing detection. The text is screened for bias, rhetorical manipulation, and misleading framing, not only for factual errors.
4. Misinformation registry cross-check. Claims are matched against known misinformation and fact-checking databases.

Independent models cross-check each other at every stage, so that no single model's blind spot becomes a blind spot for the whole system.

Human expertise stays central

Meaningful verification depends on human judgment. Journalists, fact-checkers, and analysts are essential to it, and their contributions should be recognized and fairly compensated rather than quietly absorbed into a model's output. Defacta is built to support that judgment, never to replace it.

What Defacta is not

Not a truth machine. Defacta doesn't hand down verdicts. It shows you the evidence behind each finding and how strong that evidence is, so the conclusion remains yours.

Not an AI detector. Defacta doesn't try to determine who or what wrote a text. It looks at whether the claims hold up and whether the sources behind them can be trusted.

Not an editor. Defacta doesn't rewrite or change your documents. It flags and highlights issues so you can address them in your own way.

Not a substitute for responsibility. Legal and editorial responsibility stays with you and your team, and Defacta is designed to support that responsibility rather than take it over.

What you get

• A structured report with findings, sources, confidence levels, and recommended actions
• A timestamped audit trail for every check
• Exportable reports you can share externally, without internal notes or operational metadata

Every result is traceable and source-backed, and the final decision always stays with the people responsible for it.

Where it fits

For readers. When you come across a news article, an opinion piece, a viral post, or an answer from a chatbot, Defacta helps you see beyond the surface of the text. It shows you which claims are backed by evidence, how credible the sources are, and where bias, manipulation, or misleading framing may be shaping what you read, so you can read with a deeper understanding and form your own view.

For creators. If you write social media posts, newsletters, blogs, or scripts and show notes for podcasts and videos, Defacta helps you check your facts before you publish. It points out claims that need stronger support and sources that may not hold up, so you can be more precise with your audience and protect the trust you have built with them.

For professionals and teams. Analysts, journalists, compliance teams, and communications professionals can verify a single document before it goes out, or run Defacta inside enterprise pipelines where AI agents pass work to one another. Both connect to the same audit trail. Existing workflows stay as they are, with optional API integration for teams that want verification built in.

---

For compliance teams

The evidence under your compliance process

Defacta is not a compliance platform, a case-management system, or a decision-maker. It is the layer of evidence that sits underneath your own compliance process and shows:

• What was checked
• Which sources it was checked against
• What risks were found

This gives your reviewers what they need to decide with confidence.

Where Defacta supports your obligations

Consumer protection and advertising law (US FTC Act, EU and UK consumer law) Defacta flags supported, unsupported, disputed, and risky claims before content is published, used in marketing, or relied on internally.

EU AI Act: human oversight Confidence indicators, evidence review, and a recheck workflow give reviewers something concrete to act on before AI-generated text moves into professional or operational use.

EU AI Act: transparency for AI-generated text (Article 50) Article 50 transparency obligations apply from August 2, 2026. They require disclosure when AI-generated text is published to inform the public on matters of public interest, with an exception for text that has gone through human review or editorial control, where a person or organization holds editorial responsibility. Defacta produces a timestamped record of that review, showing what was checked, against which sources, and what was found. This gives any team publishing AI-assisted content, whether in communications, public affairs, investor relations, research, or newsrooms, documented evidence of the review process the exception relies on.

EU Digital Services Act Defacta flags misinformation, manipulation, and bias to support editorial, platform, and trust-and-safety teams before publication or moderation decisions.

Source provenance Defacta records which sources were used as verification evidence, which is relevant to copyright and media-licensing accountability.

Singapore Model AI Governance Framework Defacta supports the framework's emphasis on human oversight, transparency, and accountability for AI output.

Internal governance and regulated sectors Timestamped reports with source context build a reviewable record for audits and internal controls.

What Defacta doesn't do

• Label or watermark your own AI-generated content
• Run DSA notice-and-action workflows, such as appeals, trusted-flagger processes, or transparency reporting
• Manage GDPR data-subject rights or DPIAs
• Clear copyright licenses or settle royalty disputes, since it documents source usage only as evidence

Scoped on purpose

Defacta supplies the evidence and review workflow that a compliance program needs, not the compliance program itself. It is designed to be reliable, auditable, and clearly bounded.

Defacta does not provide legal advice. Using Defacta supports, but does not by itself establish, compliance with any law or regulation.