FAQ
Frequently asked questions
How is Moyo different from DLP?
DLP watches for files and patterns leaving your network. We measure conclusions that can be assembled from what you already publish, even when no single document is “the leak.” See also idea leakage on the SenTeGuard research blog.
How is Moyo different from penetration testing?
Pen tests ask whether systems can be breached. We ask whether models can infer sensitive facts without a breach, using public (or, if authorized, in-scope internal) sources. That can be your own exposure, or what public sources already reveal about a competitor, adversary, or opponent.
Can I use Moyo on a competitor, adversary, or opponent?
Yes. Outside-in work uses lawful public sources only. Name the organization you want examined: yours, a rival, a threat actor, or a political opponent. We do not access private systems or accept confidential files. Submitting a request confirms you can ask for a public-source check of that organization.
Do you access our internal systems for an Exposure Snapshot?
No. Exposure Snapshots are outside-in, public sources only. White-box work needs written authorization and an agreed scope.
Who is the contracting party?
Moyo is a product of SenTeGuard Inc.. Contracts, invoices, and legal obligations are with SenTeGuard Inc., not a separate Moyo company.
Is there a money-back guarantee?
Yes for Exposure Data, Exposure Snapshot, and Basis Report purchases. If you are not satisfied, contact us within 14 days of delivery for a full refund. Details are in the refund policy.
Can you test our RAG system?
Yes, as a Stage 03 white-box engagement with written authorization. See Red-Teaming.
What do confidence and sensitivity mean?
Confidence reflects how reliably the inference reproduces across trials and prompting strategies. Sensitivity ranks business or mission impact if the inference is correct. Theoretical bounds are discussed in our reachability research.
Is this OSINT rebranded?
OSINT is part of the input. Moyo formalizes synthesis into evidence graphs, scores, and mitigations suitable for leadership briefings, not a loose findings dump.
How do I see what a scan would find?
Open the topic preview, then run the first Exposure Data scan free (extracted claims and structured findings in about 15 minutes). Additional scans are paid. You can also buy a Snapshot or a Basis Report packaged for presentation.
How do I read Exposure Data?
Exposure Data is extracted claims (JSON Lines) and structured findings (one JSON file), not a packaged brief. Start with the structured findings to rank and group, then use the extracted claims for evidence excerpts. Sensitivity 1–5 ranks stakes; status labels such as corroborated, contested, and outlier are not a drop list.
Can we license Moyo instead of buying reports?
Yes. SenTeGuard Inc. licenses Moyo for organizations that want to run reachability analysis under their own operational control. Licensing covers deployment model, targets, and optional white-box modules in a written agreement. Services (Exposure Data, snapshots, Basis Reports, white-box engagements) remain available if you prefer delivered work.
Does a software license include white-box access to our systems?
No by default. A license lets you operate the software. Analyzing internal corpora, RAG systems, or other non-public surfaces still requires your authorization and an agreed scope, whether you run the tool yourselves or engage us.
Still have a question? Contact us · SenTeGuard research blog.
Research
Deep dives from the SenTeGuard blog
Longer conceptual pieces on information reachability and related idea-security work are published on the SenTeGuard blog (company research site, not this product domain).
-
senteguard.com/blog
Sensitive Information Reachability: The Problem and the Solution
Core introduction to information reachability and the Moyo framing: what can be inferred from public crumbs, and how to measure it.
Open article → -
senteguard.com/blog
Defining the Limits of LLM-Reachable Intelligence with Gödelian Incompleteness
Theoretical boundaries on what models can reliably reach, useful when scoping confidence and residual risk in assessments.
Open article → -
senteguard.com/blog
The Emerging Threat of Idea Leakage
Semantic exposure without explicit disclosure: why traditional DLP misses combinatorial and proximity-based leaks.
Open article → -
senteguard.com/blog
Living with LLMs Everywhere: How Ambient LLMs Negate Security Policy
When models are embedded across tools, outbound control lists break down and inference exposure multiplies.
Open article →