Disclaimer.

1 · What the GGRI is — and isn't

The Global Geopolitical Risk Index ("GGRI") is a single 0–100 numerical indicator computed daily from a weighted combination of nine public-information components plus a small bounded random term. Full methodology is published at /methodology.html and is version-controlled.

The GGRI is a direction-of-risk indicator. It signals whether the global geopolitical risk environment is trending more or less acute on a given day. It is not:

• A prediction of any specific event, outcome, conflict, sanctions action, market move, cyber incident, or political development;
• A measure of risk to any specific person, organization, asset, location, route, or transaction;
• A substitute for trained intelligence analysis, security advance work, due diligence, or expert consultation;
• A complete picture of geopolitical risk — it captures what public OSINT can capture, which is necessarily incomplete.

2 · Not financial, legal, medical, security, or operational advice

Nothing on AtlasRisks — including the GGRI score, per-country breakdowns, Event Timelines, Watchlist alerts, the GGRI Daily brief, the Live Atlas dashboard, or any associated content — constitutes:

• Financial or investment advice. AtlasRisks is not a registered investment adviser, broker-dealer, or financial planner. Do not use the GGRI as the basis for investment, hedging, currency, or commodity decisions without independent professional advice. Past patterns do not predict future market behavior.
• Legal advice. AtlasRisks is not a law firm. Sanctions and regulatory entries surfaced in the C7 component are summaries of public actions and are not a substitute for compliance counsel.
• Medical advice. Public health items appearing in our content are journalistic summaries; consult medical professionals for health decisions including travel medicine.
• Security or operational advice. The GGRI is a country-level indicator. It is not a substitute for individual threat assessments, route surveys, advance work, executive protection planning, kidnap-and-ransom assessment, or any other operator-level diligence. Specific operations require specific professional review.
• Crisis management or emergency response advice. If you are managing an active incident, contact qualified crisis-management professionals and relevant authorities directly.

3 · Source data and credibility caveats

The GGRI ingests information from a structured registry of 40+ public OSINT sources spanning wire services, mainstream media, specialist conflict reporting, regional outlets, alternative media, government and humanitarian advisories, and (when configured) optional API integrations such as GDELT and ACLED. Each source is weighted by class — wire services and humanitarian-official feeds anchor the top of the reliability scale; alternative media and state-affiliated outlets are weighted significantly lower. Public social or open-web signals (when activated via lawful API integration) are treated as soft trend indicators, not as primary score drivers. The full source registry and weights are published in our methodology document.

Source data may be incorrect, delayed, biased, or deliberately deceptive. Specifically:

• Public news reporting can contain errors, retractions, or partial information that revises over time. We do not independently fact-check upstream reporting.
• State-affiliated sources (RT and any future additions) are flagged in the API payload (state_affiliated: true) and weighted at 0.30 — far below mainstream media. Their factual claims should be treated as the position of the affiliated state, not as verified fact.
• Alternative media sources (Bellingcat, War on the Rocks, Antiwar.com, Responsible Statecraft, The Intercept, etc.) are weighted at 0.45. They contribute corroboration value when their reporting aligns with higher-weight sources, but cannot by themselves move a country's risk score significantly.
• Geopolitical reporting is subject to active information operations and disinformation campaigns. While our pipeline includes cross-source corroboration, source-class reliability weighting, and confidence scoring, we cannot guarantee any individual signal is free of manipulation.
• Country attribution uses keyword detection plus per-feed fallback heuristics. Some attributions will be imprecise; each item exposes a country attribution confidence score (0–100) for downstream filtering.
• Where a source is unreachable from our pipeline (rate limits, geo-blocks, paywalls), the diagnostics object on each refresh records the failure and the score is computed from the remaining sources. Failures are visible in the API payload's diagnostics.errors field.
• AtlasRisks does not scrape private platforms, bypass authentication, or collect personal social media data. Any future social or open-web integration will operate exclusively via official, lawful APIs (e.g., the X paid API tier, Reddit's documented API, public Telegram channel feeds), and will be disclosed in this section.

4 · Methodology and formula caveats

The current scoring spec is v18.1. Source-class weights, scoring modifiers, and country baselines are documented in the methodology page. They are subject to revision. Score time-series across methodology versions are not directly comparable — read tier transitions and trajectory, not absolute deltas across versions.

Per-country scoring is the sum of:

• Baseline risk — manually curated 0–95 country baseline reflecting structural condition. Reviewed quarterly.
• Event severity modifier — derived from clustered events, weighted by source reliability and corroboration count.
• Trend modifier — reflects elevation above the country's quiet baseline.
• Corroboration modifier — adds weight when independent sources agree on the same event.
• Official advisory modifier — adds weight when government or humanitarian sources attribute to the country.
• Breaking modifier — preserves the v10 high-density armed-conflict signal.
• Stale data penalty — dampens the score when no recent items are available.

The result is clamped to 0–95 and bucketed into a tier (LOW / MODERATE / HIGH / CRITICAL). Each modifier is exposed in the API payload so consumers can audit which factors moved a country today. The model is intentionally explainable — we trade machine-learning sophistication for transparent decomposition that operators can read and challenge.

A separate confidence score (0–100) is published alongside each country's risk number. Confidence reflects how well-supported the score is — average source reliability, corroboration count, recency, official-source presence, and country attribution clarity. A high-risk country with low confidence is a flag to investigate, not a settled conclusion.

5 · Risk vs confidence; pseudo-precision; the noise floor

The GGRI is a 0–95 indicator. The underlying signal does not support last-decimal precision. A score of 67 today and 65 tomorrow does not reliably mean risk fell by 2 — the noise floor is wider than that, driven by ordinary daily variation in source coverage, news cycles, and event timing.

Operationally, this means:

• Read tier transitions (LOW → MODERATE → HIGH → CRITICAL), day-over-day deltas of meaningful magnitude, the confidence score, and the threat-vector breakdown — not last-decimal moves.
• A country at risk 75 with confidence 85 is a stable, well-supported signal worth acting on. A country at risk 75 with confidence 30 is a flag to investigate — likely driven by a single thinly corroborated headline.
• The delta_vs_yesterday field on each hotspot and on the global score reflects movement against yesterday's snapshot, not within-day fluctuations. Multi-day trajectory is more informative than any single-run reading.

6 · Acceptable-use boundaries

The GGRI and AtlasRisks content are not appropriate inputs for, and shall not be used as the determinative factor in:

• Individual immigration decisions, visa adjudications, asylum determinations, or border-control screening;
• Custody, parental-fitness, or family-court determinations;
• Employment screening, terminations, or denial of services to individuals;
• Insurance pricing or coverage decisions for individual policyholders;
• Targeting decisions of any kind, including kinetic, cyber, or commercial;
• Any decision affecting an individual's liberty, health, financial standing, or family integrity.

The GGRI is country-level and time-aggregated. It contains no individual data and is not designed to support individual-level adjudication.

7 · Timing, latency, and freshness

The GGRI is computed and emailed once per day at approximately 05:00 EST (10:00 UTC). Source-data freshness within that window varies by source class:

• Government primary feeds (T0): typically ≤ 1 hour latency from publication to ingest.
• Wire and major media (T1, T2): typically ≤ 30 minutes.
• Think-tank reports and analyses (T2): hours to days.
• State-affiliated sources (T3): variable, often immediate.

Real-world events between daily runs are not reflected in that day's score. Subscribers who require sub-daily freshness should not rely on the GGRI alone for active situational awareness — use the Live Atlas dashboard, which refreshes every 30 minutes (Pro and above), and primary sources.

8 · Third-party content and links

The GGRI Daily and Live Atlas link to third-party sources (news outlets, government sites, think tanks, etc.). AtlasRisks does not control, endorse, or guarantee the accuracy of third-party content. Linked content is provided for evidentiary attribution and is governed by the third party's terms.

Map tiles on the Live Atlas are provided by CartoDB and OpenStreetMap. Map projections may include territorial depictions reflecting OSM defaults; these are technical defaults of the basemap provider and do not represent AtlasRisks's position on disputed boundaries.

9 · No warranty

The Service is provided "AS IS" and "AS AVAILABLE" without warranty of any kind, express or implied, including but not limited to warranties of merchantability, fitness for a particular purpose, non-infringement, accuracy, completeness, timeliness, or freedom from interruption. AtlasRisks does not warrant that the Service will meet your requirements or that the GGRI score will be accurate, reliable, or suitable for any particular use.

10 · Limitation of liability

To the maximum extent permitted by applicable law, AtlasRisks, its operator (Cortexis Group, LLC, a Florida limited liability company), and any of its members, officers, employees, agents, or contractors shall not be liable for any direct, indirect, incidental, consequential, special, exemplary, or punitive damages — including without limitation damages for lost profits, lost data, business interruption, security incidents, decisions made in reliance on the GGRI, or harm arising from third-party source content — even if advised of the possibility of such damages. Our aggregate liability for any claim arising out of or relating to the Service shall not exceed the fees paid by the claimant to AtlasRisks in the twelve (12) months preceding the claim.

Some jurisdictions do not allow the exclusion or limitation of certain damages; in such jurisdictions our liability is limited to the maximum extent permitted by law.

11 · Attribution and republication

Subscribers may quote, cite, and reference the GGRI in internal briefings, client deliverables (when relevant to your work and properly attributed), and academic research. Required attribution: "Source: AtlasRisks GGRI, [date]."

The following are not permitted without a written license:

• Republication of the GGRI score or daily brief in publicly distributed media;
• Resale, sublicense, or redistribution of GGRI data to non-subscribers;
• Use of the GGRI as a feed into a competing risk-rating product;
• Reverse-engineering of the methodology to clone the score for commercial use;
• Bulk scraping of the dashboard, API, or daily emails beyond the rate limits applicable to your tier.

Enterprise tier includes redistribution rights for use within the licensee's organization. Contact legal@atlasrisks.com for licensing.

12 · Changes to the formula and weights

Source-class weights, scoring modifiers, country baselines, and the source registry are versioned (current scoring spec: v18.1). Material changes are announced via the methodology page changelog and to active subscribers. Score time-series across versions are not directly comparable; subscribers should be aware that a methodology bump can produce a step-change in score independent of any change in underlying conditions.

Backtests and historical comparisons are performed on the methodology version current at the time of the run, unless explicitly versioned otherwise.

13 · Regulatory note

AtlasRisks is not registered as an investment adviser under the U.S. Investment Advisers Act of 1940, the U.K. Financial Services and Markets Act 2000, the EU MiFID II framework, or any analogous regulation. We are not a credit-rating agency under Regulation (EC) 1060/2009 or the SEC's regulation of NRSROs. The GGRI is not a credit rating, sovereign risk rating, or regulatory rating of any kind.

If you are operating in a regulated industry (banking, insurance, asset management, pensions, etc.) and considering use of the GGRI as a regulatory input, you should obtain compliance counsel before doing so.

14 · Contact for clarifications

For questions about this disclaimer or the appropriate use of GGRI data:

• General: hello@atlasrisks.com
• Legal: legal@atlasrisks.com
• Methodology questions: methodology@atlasrisks.com

— AtlasRisks Disclaimer · scoring spec v18.1 · revised 2026-05-16