Professional intelligence is traditionally split into collection disciplines — the INTs: OSINT (open sources), HUMINT (human sources), SIGINT (signals), GEOINT/IMINT (imagery), and MASINT (physical signatures). The striking shift of the last few years is democratization: open-source intelligence now makes up an estimated 80–90% of Western intelligence activity, and commercial tools have pulled several disciplines into public reach — 30 cm satellite imagery, open ship and aircraft tracking, image forensics, link analysis, and facial recognition. RAND-cited work suggests small OSINT teams can reach 70–90% of the analytic value of classified collection at roughly 2% of the cost. But the overlap is not total: HUMINT (intentions and secret plans), MASINT (nuclear and chemical signatures), and classified all-source fusion remain agency advantages — and there is a real access gap. This guide maps each discipline to the open tools that now approximate it.
The intelligence disciplines, in one minute
Professional intelligence organizes collection into disciplines known as the INTs, for their shared suffix. Per Mark Lowenthal and Robert Clark's standard text and the U.S. Office of the Director of National Intelligence, the classic five are:
- OSINT — open-source intelligence, from publicly available information: media, public records, social platforms, commercial databases, satellite imagery, and more.
- HUMINT — human intelligence, from people: interviews, debriefings, diplomatic reporting, and clandestine sources. The oldest discipline, and until the late-20th-century technical revolution, the primary one.
- SIGINT — signals intelligence, from intercepted transmissions; includes COMINT (communications) and ELINT (electronic/radar emissions). The NSA is the U.S. lead.
- GEOINT / IMINT — geospatial and imagery intelligence, from satellite and aerial imagery, produced by the NRO and NGA.
- MASINT — measurement and signature intelligence, from physical signatures: nuclear, radar, acoustic, chemical, thermal. The most arcane discipline, held by only a few nations.
Newer frameworks add CYBER and sometimes FININT (financial intelligence). Crucially, no single INT sees the whole picture; agencies fuse them into all-source intelligence, where each discipline corroborates and completes the others. That fusion is the thread to keep in mind — it's where the agency advantage ultimately lives.
How much intelligence really comes from open sources?
More than most people assume. A widely cited 2023 systematic review by Ghioni, Taddeo, and Floridi (in AI & Society) estimates that OSINT accounts for 80 to 90 percent of the intelligence activity of Western law enforcement and intelligence services. The lineage is older than the internet: in 1947, Allen Dulles told the Senate that overt, above-board sources could supply over 80 percent of the information needed to guide national policy — a figure later echoed by senior officials including a former director of the Defense Intelligence Agency, and now put by some practitioners closer to 90–95 percent.
The resource math is what makes this consequential. Work cited by RAND and in congressional testimony suggests small OSINT cells can generate 70–90 percent of the analytic value of classified collection while operating at roughly two percent of the cost. As CSIS put it in its analysis of the democratization of intelligence, open sources often surpass classified information in speed, scale, clarity, usability, and cost. The U.S. Intelligence Community formalized the shift in its OSINT Strategy 2024–2026, which elevated open source to a first-tier discipline; in 2026 the DIA went further, merging its OSINT and media-exploitation functions.
The democratization: professional capabilities, now open
Here is the heart of it — a discipline-by-discipline map of the classic professional capability and the open or commercial equivalent that now approximates it. Where Max Intel offers a browser-based version, it's linked.
| Discipline (INT) | Classic professional capability | Open / commercial equivalent (2026) |
|---|---|---|
| GEOINT / IMINT | NRO/NGA reconnaissance satellites; agency imagery analysts | 30 cm commercial imagery (Maxar/Vantor, Planet, BlackSky, Airbus); free 10 m Copernicus Sentinel Hub; AI geolocation → Geolocation, EXIF Viewer |
| SIGINT (adjacent) | NSA interception; classified COMINT/ELINT | Open AIS (ships) & ADS-B (aircraft) tracking — MarineTraffic, ADS-B Exchange, FlightRadar24 — at granularity approaching signals collection; commercial RF scanners; data brokers → Network, IP |
| Digital forensics | Agency forensic laboratories | Error-level analysis, metadata, manipulation detection → Photo Forensics, Steganalysis |
| Link analysis | Palantir Gotham; i2 Analyst's Notebook | Maltego, ShadowDragon → Entity Graph |
| Facial recognition | Classified biometric matching systems | PimEyes and similar face-search services → Face Finder |
| Identity resolution | Agency person and travel databases | Username, email, and people-search OSINT → Person, Username, Email |
| HUMINT | Clandestine human sources; espionage | Not democratized — overt interviewing and social-media intelligence only partially overlap |
| MASINT | Classified nuclear, chemical, and radar signature sensors | Not democratized — remains an agency-only capability |
Two examples show how far the imagery side has moved. Commercial satellite imagery from Maxar (now Vantor) and Planet Labs supplied much of the public's view of Russian troop movements before and during the invasion of Ukraine, and in the 2026 US–Iran conflict, observers watched strikes unfold through near-real-time commercial imagery — coverage once reserved for the top tier of government. On the signals-adjacent side, the "privatization of SIGINT" is real: commercial radio scanners have picked up military chatter, and open flight- and vessel-tracking routinely surfaces movements that used to require an intercept.
Where OSINT overlaps professional intelligence
The overlap is genuine and growing, especially in imagery, geolocation, and movement tracking, where open data can match classified collection in speed and often beats it to publication. Bellingcat's investigations have repeatedly demonstrated open-source work rivaling — and occasionally outpacing — official assessments. This is why agencies increasingly practice what former CIA deputy director Carmen Medina calls intelligence integration: point scarce classified collection at the gaps open sources can't fill, and use public analysis for context, corroboration, and amplification. Ukraine's wartime intelligence, blending government feeds with crowdsourced reporting and commercial satellite analysis, is the textbook case of open and classified sources producing more together than either alone.
What OSINT still can't do
It's worth being precise about the limits, because overclaiming is its own failure mode. Three boundaries persist:
- HUMINT. Public data can reveal what happened and where, but not reliably why, or what a decision-maker intends to do next. Recruiting and running human sources for intentions, secret plans, and undocumented decisions has no open-source substitute.
- MASINT and classified sensors. Detecting a nuclear test signature or characterizing an unknown weapon system needs specialized sensors that remain the preserve of a few states.
- Classified all-source fusion. Even where imagery is democratized, an access gap remains: a President's Daily Brief is an all-source product combining commercial imagery with signals, human, and technical intelligence the public never sees. Providers can also throttle access — both Vantor and Planet restricted imagery flow during the 2026 Iran campaign.
There's also a serious scholarly counterpoint worth taking on directly. A 2025 analysis in the European Journal of International Security argues that the rise of OSINT is not truly "democratising intelligence": high-end OSINT demands scarce subject-matter expertise, source familiarity, and increasingly expensive AI tooling, and even skilled experts struggle to interpret and contextualize open data. And OSINT carries its own failure modes — disinformation and manipulated media, stripped context, transponders switched off to "go dark," and what one analysis calls the illusion of accuracy, where unverified data circulates as fact. Access to tools is not the same as the judgment to use them well.
What this means for you
The practical upshot: an enormous amount of what used to require an agency is now a browser tab. Max Intel exists in exactly this democratized layer — its tools map onto the open equivalents in the table above, from geolocation and imagery metadata to link analysis and facial search. The 2026 frontier is automating that work: see the AI-agent OSINT & MCP guide for how AI agents now chain these capabilities autonomously. But democratized capability comes with democratized responsibility. Use these tools for authorized research, journalism, and security work — never for surveillance, harassment, or intimidation — verify every finding against its source, and remember that the gap between OSINT and professional intelligence is now less about access and more about tradecraft, verification, and the judgment to know what a finding actually means.
Sources and further reading
Office of the Director of National Intelligence (dni.gov, "What is Intelligence"); Lowenthal & Clark, The Five Disciplines of Intelligence Collection; the Ghioni, Taddeo & Floridi 2023 review in AI & Society; CSIS, "Responding to OSINT: U.S. Strategy and the Democratization of Intelligence"; the RAND-cited analysis via Small Wars Journal; Lawfare on an IC OSINT agency; the European Journal of International Security (Cambridge) on OSINT's limits; and Bellingcat. Capability figures are directional and change quickly — verify current commercial imagery resolution, tool availability, and legal constraints before relying on any of them.