Cross-domain intelligence: the value proposition
The most consequential risks rarely respect domain boundaries. A sanctions development that simultaneously affects payments corridors, AI governance, democratic integrity, and conflict escalation is invisible to any monitor that watches only one of those domains. By the time each domain-specific monitor has registered the development independently, the pattern has already become obvious — and the analytical window has closed.
Signal exists to surface these cross-domain patterns before they become obvious. It is built on a single insight: the most valuable intelligence sits at the intersection of domains, not within them.
Signal's analytical position: Cross-domain visibility is not just more coverage — it produces qualitatively different intelligence. Some patterns are structurally invisible to single-domain monitors. A regulatory pivot that looks like a policy footnote in one domain looks like a regime transition when it coincides with stress signals in three others. Signal exists to surface that difference.
This methodology document describes what Signal tracks, which source monitors it draws from, how cross-domain patterns are identified and classified, what a cross-domain signal looks like in practice, and what Signal does not do.
What Signal tracks
Signal does not track individual domains. The seven source monitors each do that, with their own dedicated methodology and source bases. Signal tracks the intersection — the structural relationships between domains, and the patterns that only become legible when multiple domains are read simultaneously.
Specifically, Signal tracks three categories of inter-domain pattern:
- Convergence — when independent monitors, drawing on separate source bases, register the same risk or structural feature without coordination. Convergence elevates analytical confidence beyond what any single monitor can achieve alone.
- Divergence — when monitors that would normally move together instead pull apart. Divergence between historically correlated domains is itself a signal: it indicates mispricing, an information asymmetry, or an early-stage regime transition that has not yet registered across the full system.
- Cascade — when a development in one domain demonstrably triggers or accelerates developments in others, producing a causal chain that travels across the analytical map. Cascade patterns are the hardest to detect in real time and the highest-value findings Signal publishes.
Signal does not summarise the source monitors. It does not produce a digest of what each monitor found that week. Its function is pattern detection — identifying the structural relationships between domains that are invisible to any monitor operating alone.
Source monitors
Signal draws on all seven commons monitors published by Asymmetric Intelligence at asym-intel.info, plus the commercial intelligence feeds from the Asymmetric Intelligence network. Each monitor operates with its own methodology, source hierarchy, and publication cadence — detailed in its own methodology documentation. Signal reads the outputs of all contributing monitors as a corpus and applies cross-domain analysis on top of that primary work.
| Monitor | Full name | Primary domain | Methodology |
|---|---|---|---|
| WDM | World Democracy Monitor | Democratic backsliding, institutional integrity, election engineering, civil society | methodology ↗ |
| GMM | Global Macro Monitor | Macro-financial stress, sovereign risk, supply chain disruption, capital flows | methodology ↗ |
| FCW | FIMI & Cognitive Warfare Monitor | Foreign information manipulation, influence operations, narrative attribution | methodology ↗ |
| ESA | European Strategic Autonomy Monitor | European-theatre geopolitics, hybrid threats, strategic dependency, defence integration | methodology ↗ |
| AIM | Artificial Intelligence Monitor | AI capability development, governance, regulation, deployment risk, standards | methodology ↗ |
| ERM | Environmental Risks Monitor | Planetary boundaries, climate tipping points, physical risk, climate-security nexus | methodology ↗ |
| SCEM | Strategic Conflict & Escalation Monitor | Conflict escalation dynamics, theatre assessment, military signalling, strategic risk | methodology ↗ |
Signal does not replace or summarise these monitors. Subscribers seeking single-domain depth should read the monitors directly — all seven are publicly accessible at asym-intel.info. Signal's distinct value is in cross-domain correlation: the patterns that only become visible when all contributing monitors are read together.
Each monitor publishes on a staggered weekly schedule. This sequencing is analytically significant: monitors publishing later in the week can reference findings from earlier publications, allowing cross-domain signals to emerge within the same analytical cycle rather than only across cycles.
Pattern detection framework
Signal applies three standing pattern types to the cross-monitor corpus each week. Each type describes a different kind of inter-domain relationship and warrants different analytical treatment.
Convergence is the primary mechanism by which cross-domain analysis produces higher-confidence intelligence than single-domain analysis. When two or more monitors — each using independent source bases — identify the same structural risk or causal pattern without coordination, the probability that the pattern reflects a real-world phenomenon is meaningfully higher than if only one monitor had observed it. Signal distinguishes hard convergence (same specific actor or event identified from separate evidentiary trails) from structural convergence (same type of pattern identified independently across domains), and temporal convergence (independent monitors registering elevated stress in overlapping time windows).
Divergence between historically correlated monitors is analytically as important as convergence. When financial signals are escalating while market pricing remains stable, or when environmental stress indicators are rising while macro signals remain benign, the divergence itself is the finding. It indicates one of three things: a lead-lag relationship (one monitor is ahead of the other), an information asymmetry (one monitor's source base has a systematic gap the other has not yet registered), or genuine decorrelation between domains (often a precursor to a regime transition). Signal tracks divergence systematically across all monitor pairs, not opportunistically. Sustained divergence over multiple publication cycles is escalated to a featured finding.
Cascade patterns occur when a development in one domain demonstrably triggers or accelerates developments in others — producing a traceable causal chain that crosses domain boundaries. A military escalation that simultaneously drives economic coercion, FIMI campaign intensity, and domestic democratic pressure is a cascade. So is an AI governance decision that reshapes both the competitive landscape for AI capability and the regulatory environment for information operations. Cascade identification requires tracing causation, not merely observing coincidence. Signal requires directional linkage — a plausible mechanism by which the originating development affects subsequent domains — before classifying a pattern as a cascade.
Materiality threshold
Not every topical overlap between monitors constitutes a signal. The threshold for Signal to surface a cross-domain pattern is materiality: the inter-domain relationship must either change the analytical assessment of a domain, indicate a risk that would be missed by single-domain monitoring, or qualify as an early indicator of a macro transition. Informational overlap without analytical consequence is not published.
Source independence verification
The analytical value of convergence depends entirely on the independence of the contributing monitors. Before treating cross-monitor agreement as confidence-elevating, Signal verifies that convergent findings rest on genuinely different source bases. If two monitors are both relying on the same underlying primary source or the same secondary reporting outlet, that agreement is source-contaminated and treated as a single observation. Independence is a prerequisite, not an assumption.
Source hierarchy
Signal inherits the source-tier methodology applied across all Asymmetric Intelligence monitors. It does not have a separate source base — its sources are the outputs of the contributing monitors, which in turn apply their own documented source hierarchies. Signal adds a synthesis tier on top of this inherited structure.
| Tier | Type | Role in Signal |
|---|---|---|
| Synthesis | Signal cross-domain analysis | Signal's own layer — identifies convergence, divergence, and cascade patterns across the contributing monitor corpus. This tier does not generate new primary intelligence; it identifies structural relationships between domain-specific findings. |
| T1 | Primary source documentation | Official records, regulatory filings, legal proceedings, primary data — sourced and verified by contributing monitors. Signal inherits T1 attributions from the monitor that established them. |
| T2 | Institutional reporting | Established investigative journalism, think-tank primary research, academic publications. High evidentiary weight; independently verified by rigorous process within the contributing monitor. |
| T3 | Analytical synthesis | Expert commentary, policy analysis, structured secondary reporting. Moderate weight; used to contextualise, not to establish findings. |
| T4 | Indicator-level signals | Open-source signals, market data, public statements with limited corroboration. Useful for pattern recognition and early-stage flagging; insufficient on their own for published cross-domain claims. |
The inheritance model
Each contributing monitor maintains its own source tier assignments for every finding. When Signal draws on a monitor's output, it inherits those tier assignments. A finding sourced to T1 by WDM arrives in Signal's corpus as a T1 finding. Signal does not re-evaluate or downgrade source attributions made by contributing monitors — it applies those attributions within its cross-domain analysis.
Signal adds one rule not present in contributing monitors: cross-domain convergence can raise the effective confidence of a cross-domain finding. A finding rated "Assessed" based on a single monitor's T3 sources may be elevated to "High" confidence if the same structural pattern is independently observed at T2 or above in a second monitor. This elevation is applied conservatively and requires genuine independence of observation.
Full source methodology for each contributing monitor is available at asym-intel.info.
Illustrative examples
The following are hypothetical examples constructed to illustrate what a cross-domain signal looks like in practice. They do not reflect actual findings from any Signal publication. They are provided to help readers understand the type of pattern Signal is designed to surface.
A major sanctions designation targeting a technology conglomerate would be registered by GMM as a macro-financial event — capital flow disruption, currency pressure in exposed markets. FCW would independently flag an immediate intensification of FIMI activity attributed to affiliated networks, consistent with a pattern of narrative counter-mobilisation following sanctions pressure. AIM would note that the designated entity had been a primary supplier to AI infrastructure across several jurisdictions, triggering regulatory uncertainty in AI governance frameworks dependent on that supply chain. WDM would observe that the sanctioned entity's domestic political connections accelerated legislation targeting opposition civil society groups. No single monitor would detect the full pattern. Signal would surface the cascade: a single designations event producing simultaneous and causally related stress across macro, cognitive warfare, AI governance, and democratic integrity domains.
ERM might register sustained acceleration in a key physical risk indicator — a tipping-system threshold approaching faster than modelled — while GMM simultaneously registers stability or even improvement in commodity market pricing for the affected regions. Under normal conditions, physical risk escalation at ERM would be expected to correlate with macro stress signals at GMM. When the two instead diverge, Signal treats this as an anomaly requiring explanation. The divergence could indicate that financial markets are systematically mispricing physical risk, that GMM's source base has a gap in physical-risk coverage, or that a structural decoupling between the two domains is underway. Each possibility is analytically important. Signal surfaces the divergence and its potential interpretations; it does not resolve the divergence on behalf of either monitor.
FCW might document a FIMI infrastructure network using a specific combination of jurisdictions — a particular beneficial ownership arrangement spanning three low-disclosure territories — as its distribution and funding layer. GMM might independently document a sanctions-evasion financial structure using a materially similar jurisdictional architecture, identified through a separate investigative trail. ESA might flag a covert procurement operation for dual-use technology that exploits the same jurisdictional combination. None of the three monitors would have grounds to connect these observations — they involve different actors, different objectives, and different analytical domains. Signal would identify the structural convergence: the same jurisdictional infrastructure pattern appearing independently across cognitive warfare, illicit finance, and procurement evasion. The convergence does not establish that the same actors are responsible — it establishes that the architecture is a shared enabling layer and warrants cross-domain investigation.
Limitations
Signal is explicit about what it does not do and where its methodology has edges.
- Signal does not replace single-domain depth. The contributing monitors each provide substantially more detail, nuance, and domain-specific context than Signal can incorporate in a cross-domain synthesis. Readers who need the full analytical picture for a specific domain should read the relevant monitor directly.
- Signal is a weekly synthesis, not a real-time feed. Cross-domain patterns require the outputs of contributing monitors to be read in aggregate. Most contributing monitors publish weekly. Signal's synthesis is therefore weekly. Developments between publication cycles — including rapidly evolving situations — will not appear in Signal until the following publication cycle unless they are flagged in a contributing monitor's interim communication.
- Cross-domain confidence elevation is not certainty. When Signal elevates the confidence of a finding based on convergence across multiple monitors, it is making a probabilistic judgment. Convergence raises the probability that a pattern reflects a real-world phenomenon; it does not confirm it. Signal's premium confidence designation ("Cross-validated") still requires readers to apply their own judgment about whether the converging evidence base is genuinely independent.
- Signal does not attribute intent. Identifying structural convergence between networks, or establishing that a cascade has propagated across domains, does not on its own establish coordination, intent, or shared purpose among the actors involved. Attribution claims beyond structural observation require T1 evidence from contributing monitors.
- Coverage gaps in contributing monitors are inherited. Signal's cross-domain analysis is bounded by the coverage of contributing monitors. Domains, geographies, or actors not within any contributing monitor's scope are not visible to Signal. Signal does not compensate for these gaps; it inherits them.
- Pattern detection is not prediction. Signal identifies patterns that have already emerged across domains. It does not forecast which cross-domain patterns will emerge in future cycles. Trend direction assessments are published where the evidence supports them, but Signal does not publish probabilistic forecasts of future cross-domain events.
The analytical ecosystem
Signal is the cross-domain synthesis layer of the Asymmetric Intelligence network. It is structurally dependent on, and structurally inseparable from, the seven commons monitors it draws from. Those monitors are independently valuable as single-domain intelligence products. Signal is valuable precisely because they exist — the quality of cross-domain synthesis is bounded by the quality of the domain-specific intelligence underneath it.
The relationship is bidirectional. Cross-domain patterns identified in Signal frequently generate cross-monitor flags that are incorporated into subsequent publications by individual contributing monitors. A cascade pattern surfaced in Signal may prompt WDM to watch for democratic integrity implications of a development it had previously assessed as confined to the conflict domain. A divergence anomaly flagged in Signal may lead GMM to investigate whether its macro source base has a physical risk gap that ERM has not yet flagged through normal channels.
The Asymmetric Intelligence network also publishes a live cross-monitor network graph at asym-intel.info, visualising the active connections between all contributing monitors as a force-directed map. This graph reflects the same inter-domain relationships that Signal analyses in its weekly synthesis, rendered as a navigable network for readers who want to explore the connection landscape directly.
Signal subscribers are encouraged to read the contributing monitors directly — all seven are publicly accessible, free of charge, at asym-intel.info. Signal and the commons monitors are complementary: the monitors provide domain depth; Signal provides the cross-domain view that the monitors alone cannot.
Editor
Editor: Peter Howitt
Signal is published by Asymmetric Intelligence. The methodology described in this document governs all Signal publications. Questions about methodology should be directed to the editorial contact at a-i.gi.
Version history
This table records public changes to the Signal methodology. Internal operational changes — including adjustments to analysis processes, source updates, or workflow sequencing — are not listed here.
| Version | Date | Change |
|---|---|---|
| 1.1 | Apr 2026 | Methodology page revised. Added cascade pattern type to detection framework. Expanded source monitors table to include all seven commons monitors with links to individual methodology pages. Restructured source hierarchy to make inheritance model explicit. Added illustrative examples section. Clarified Signal does not produce real-time output and does not attribute intent. |
| 1.0 | Jan 2026 | Initial public methodology page published for signal.gi launch. |