In my previous article on risk analysis, I state that the goal of conducting rapid and efficient operation in the information environment (OIE) was not to move fast by cutting corners. It is to move fast by eliminating dead time, the hours spent aggregating data, formatting packets, routing emails, and waiting for cold reads. The methodology, the rigor, and the human decision points stay the same. What changes is the time between them. The result: OIE trigger to tasking in two to eight hours, with full legal review, simulation execution (a new step that replaces murder boards), risk assessment, stakeholder coordination, and command approval preserved at every stage.
Five root causes account for most of the delay in cognitive operations staffing. 1 – Ambiguous ownership means cognitive operations fall between organizational seams, not quite intelligence, not quite operations, not quite public affairs, and when no one is clearly responsible, everyone waits. 2 – Serial processing moves packets sequentially through legal, public affairs, intelligence, and command, with each handoff introducing delay. 3 – Undefined standards cause reviewers to either over-review or rubber-stamp. 4 – Risk aversion without a risk framework causes every potential concern to become a reason to delay. And 5 – incomplete (or inconsistent) packets arrive missing information, triggering requests that add hours or even days. Every one of these failure modes is a function of process design, not operational necessity.
The IRIS-AI-Approval Automation Triad is the platform architecture that addresses all five. The Interactive Realtime Information System (IRIS) is the data foundation, a curated repository of historical behavioral and operational data with live ingest across social media, media intelligence, geospatial, demographic, psychographic, survey, and event data, refreshed on 15-minute cycles. IRIS eliminates the blank-page problem: every new Cognitive Target Nomination Packet (CTNP)/ Strawman Profile/ Target Audience Analysis begins with a pre-populated draft drawn from institutional memory. The AI layer performs pattern recognition at scale, sentiment analysis, network mapping, trend detection, anomaly alerts, and auto-populates all CTNP fields, transforming every phase from a building task into a refinement task. The Approval Automation layer routes packets to all reviewers simultaneously, enforces deadlines, maintains version control, generates immutable digital signatures, and produces post-approval tasking orders. A two-hour CTNP is as auditable as one that took two weeks.
At scale, an Influence Decision Compression Framework, underpinned by the IRIS-AI-Approval Automation Triad, is necessary to delineate how all Information Forces Capabilities compress their respective staffing cycles and build a common CTNP for consistent staffing. In my mind, from T+0 to T+30 minutes, AI performs continuous monitoring across the Kill Web’s five aspects, detects anomalies, and generates an automated alert with preliminary context based on command desired end-states. From T+30 minutes to T+2 hours, AI auto-populates the CTNP Minimum Viable Packet (MVP), target profile, HPEM position, latitude assessment, proposed cognitive effect, drafts Lines of Persuasion, baseline measures, data confidence scores, and legal flags, so the packet author refines rather than builds. From T+2 to T+4 hours, all reviewers receive pre-screened packets simultaneously for review and approval: legal review with possible/known restrictions pre-flagged, intelligence validation with source reliability pre-assessed, an initial IGL analysis complete, monte-carlo analysis assessments with clear variable statements, PSYOP feasibility with HPEM consistency pre-checked, proposed public affairs guidance, and an eight-factor risk analysis pre-scored. From T+4 to T+5 hours, AI generates a finalized dissemination plan with optimal channels, messenger profiles, timing windows, and cadence recommendations. From T+5 to T+8 hours, the commander receives a complete, pre-vetted, risk-scored CTNP with three decision options and their implications. The decision is faster because the preparation is better, consistent, and ideally integrated into the Joint Targeting Cycle.
The proper role of AI in this framework is friction reduction, not oracle consultation. AI does pattern recognition, consistency, option generation, options testing, source citation with clearly identified as well as verifiable rationales, and routine automation exceptionally well. It does not do contextual judgment, cultural nuance, ethical reasoning, or accountability. At no point does AI make decisions. AI generates inputs for human decision-making. No AI-generated message reaches a target audience without human approval. “Because the model said so” is never an acceptable justification for operational decisions. The human-AI partnership model is straightforward: AI accelerates, humans validate.
The platform also enables scale. Without the triad, a cognitive operations team can manage three to five concurrent CTNPs. With it, each member of a team can now manage ten to twenty concurrent operations, not by working harder, but by automating the tasks that require effort without requiring judgment. The hundredth CTNP is faster than the first, not because the methodology changed, but because the platform learned from the ninety-nine that came before. The methodology exists. The technology exists. Speed and legitimacy are not opposites. Rigor and scale are not incompatible. The AI-enabled-CTNP framework is the operational bridge that enables speed, scale, rigor and legitimacy. Imagine where we will be with a system like this and where our adversaries are already going, can you afford to wait?
