# Talking Tradecraft Part 1: The Executive Perspective on Investigative AI > In Part 1 of the Talking Tradecraft series, Babel Street leaders discuss investigative tradecraft, agentic AI, human judgment, and the future of intelligence. **Published:** Sep 3, 2026 **Topics:** Agentic Risk Intelligence --- > **What Is Investigative Tradecraft?** > > Investigative tradecraft is the disciplined methodology investigators and intelligence analysts use to collect information, evaluate evidence, validate sources, and reach defensible conclusions. In Part 1 of the Talking Tradecraft series, Babel Street's executive leaders explain why these principles remain essential as organizations adopt agentic AI and AI-assisted investigative workflows. Since Babel Street was founded, our focus has been on enabling intelligence analysts and investigators to find and analyze the data they need to be effective and make insightful decisions. That mission hasn’t changed, but the speed and scale at which these experts operate has, so we’ve continued to change too. With the recent release of [Babel Street Insights Investigator](https://www.babelstreet.com/insights-investigator-agentic-ai-risk-intelligence), an agentic AI capability that incorporates tradecraft, we began examining what tradecraft actually means, why it matters more than ever in the age of AI, and how organizations should think about the future of investigations. For the first installment of our Talking Tradecraft series, we sat down with several Babel Street leaders: Benji Hutchinson, CEO; John Larson, President and Chief AI Officer; and Pat Butler, Executive VP of Strategy, to discuss their views on tradecraft. Their answers revealed a consistent theme — technology changes, but disciplined investigative thinking remains the foundation of intelligence. Here’s what else we learned. ## What investigative tradecraft really means Fifteen years ago, the challenge was often finding the right data. Today, the challenge is finding meaning within an overwhelming flood of it. The explosion of open-source information, combined with advances in artificial intelligence, has fundamentally altered the investigative landscape. Yet amid this change, one thing has remained constant: the importance of tradecraft. > _Experienced investigators don't simply know where to look. They understand how to think._ For Pat Butler, tradecraft is best understood as accumulated decision-making expertise. He said, “Over time, people develop expertise and they've seen enough patterns that they develop a mental decision tree. Tradecraft is understanding that decision tree with greater and greater fidelity. The more expert a person is, the more branches on that decision tree they have come to understand.” For John Larson, tradecraft is revealed through examples. A launch vehicle expert who can infer capabilities from subtle visual indicators is demonstrating tradecraft. An investigator who can start with a social media profile and methodically connect it to related digital identities, contact information, and physical-world indicators is applying tradecraft. In each case, specialized knowledge is transformed into a repeatable investigative process. Benji Hutchinson approaches tradecraft from a mission perspective. The specific methods may vary between a commercial risk team, a military intelligence analyst, or a national security investigator, but the underlying principle remains the same. He said, “Tradecraft could mean different things, but it always goes back to their mission. There are underlying sources and methods that are employed to help effectively execute a mission. That's the way I think about tradecraft.” What unites these perspectives is the recognition that tradecraft is not a tool. It is judgment applied through process. ## The shift from data scarcity to data saturation in intelligence analysis The rise of [open-source intelligence](https://www.babelstreet.com/osint-and-threat-intelligence-solutions) has transformed the investigative environment. Modern analysts have access to more publicly available information than any previous generation. Social media platforms, messaging applications, online forums, public records, commercial databases, dark web sources, and global news coverage create unprecedented visibility into people, organizations, events, and emerging risks. This abundance creates opportunity, but it also introduces new challenges. When information was scarce, collection itself often created value. Today, collecting more information does not necessarily produce better intelligence. In many cases, it simply creates more noise. According to our executives, one of the most important shifts in modern investigations is the growing importance of determining which information matters. The challenge is in identifying signal, understanding context, and producing findings that meaningfully inform decisions. This shift has elevated the importance of analytical rigor. The ability to ask better questions, formulate better hypotheses, and connect disparate information often matters more than the volume of information available. ## Where AI falls short in investigative tradecraft The market narrative around AI tends to focus on speed, but according to our team, a fast answer has little value if analysts can’t explain where it came from. A recurring concern involved the tendency of many AI systems to produce persuasive-looking conclusions without providing clear evidence of how those conclusions were reached. Hallucinations remain a challenge, but so do issues of traceability, provenance, and repeatability. Experienced investigators naturally question assumptions, consider alternative explanations, and actively look for reasons they may be wrong. AI systems do not consistently do this on their own. The danger with relying solely on AI is that it can create unwarranted confidence in answers that have not been sufficiently challenged. Strong tradecraft acts as a corrective mechanism. It forces investigators to test assumptions, validate sources, and examine competing explanations rather than accepting the first plausible conclusion. ## What investigative tradecraft must become in the agentic AI era As organizations begin adopting [agentic AI](https://www.babelstreet.com/solutions/strategic-threat-intelligence) systems capable of planning, executing, and orchestrating multi-step workflows, the role of tradecraft becomes even more important. According to Benji, “AI is going to create a world where we have more semi-autonomous agents working on behalf of case officers or operators. We're going to have to use agents and agentic architectures to help find the signal through the noise.” Going forward agents will perform substantial portions of collection, correlation, research, and workflow execution. But successful systems won’t simply automate tasks. They will encode the logic, discipline, and decision frameworks developed by experienced investigators. In practical terms, that means agentic systems must understand how to decompose problems, pursue multiple avenues of inquiry, validate findings, and identify potential gaps in evidence before presenting conclusions. In other words, agents must operate according to proven investigative principles. The standard for these systems should not be a fast or plausible answer. It should be a decision-ready finding: citable, traceable, corroborated, clear about confidence, explicit about competing explanations, and honest about gaps in the evidence. Tradecraft must also govern the boundaries of agentic action. As agents move from generating answers to planning and acting, organizations need clear rules for what those agents can access, what actions they can take, where human approval is required, and how every action is logged and audited. ## Why human judgment remains essential in AI-assisted investigations The rise of agentic AI has prompted questions about the future role of analysts. The answer from Babel Street's leadership was consistent: AI should reduce manual involvement, not diminish human judgment. As routine collection and research activities become increasingly automated, analysts should spend less time gathering information by hand and more time providing oversight, context, prioritization, and accountability. John's view is that the role of the human analyst does not disappear as agentic systems become more capable. Instead, human judgment remains essential because someone must determine whether the output is credible, sufficiently supported, and fit for purpose. He said, “At the end of the day, humans must be at the center of AI — to look at, to evaluate, to assess, to corroborate, to push back on the agents and tell them this isn't the right answer. The whole ecosystem must be engineered to ensure that that humans end up in control of the agents.” This matters because accountability cannot be delegated. Organizations may automate workflows, but they still need humans responsible for conclusions, decisions, and outcomes. ## From government intelligence to commercial risk investigations Many tradecraft concepts originated in government and intelligence environments, but the executives interviewed believe they have become equally relevant in commercial settings. Corporate security teams, fraud investigators, compliance professionals, [supply chain risk managers](https://www.babelstreet.com/supplier-vetting-for-supply-chain-risk-management), and executive protection programs increasingly face the same fundamental challenge confronting intelligence agencies: too much information and not enough clarity. As regulatory scrutiny increases and organizational decisions become more dependent on data-driven assessments, concepts such as defensibility, traceability, and source validation become business requirements rather than intelligence-specific concerns. The organizations that perform best won’t necessarily have access to the most data. What they will have are the strongest processes for transforming information into trusted intelligence. ## The future belongs to organizations that institutionalize investigative tradecraft The next generation of risk intelligence will be defined by who applies AI most effectively. The organizations that gain a lasting advantage will combine human expertise, proven investigative methodologies, trusted data, and agentic systems capable of executing at scale. They will build workflows that prioritize transparency over black-box automation, velocity with confidence over unchecked speed, and defensible outcomes over superficial productivity gains. The opportunity is not to choose between speed and confidence, rather it is to achieve velocity with transparency and defensibility, so organizations can move faster while still understanding how conclusions were reached and why they can be trusted. Tradecraft has always been about turning information into action. In the agentic era, the organizations that succeed will be those that turn tradecraft itself into a scalable capability—one that empowers analysts, guides AI, and produces intelligence leaders can trust. > **Coming soon** > > [In part 2 of our Talking Tradecraft series](https://www.babelstreet.com/blog/talking-tradecraft-part-2-tradecraft-at-scale), learn how Babel Street incorporates tradecraft into Insights Investigator workflows. > > [In part 3](https://www.babelstreet.com/blog/talking-tradecraft-part-3-how-investigators-think), we speak with our practitioners to learn the rules they follow for sound investigative tradecraft. ## Frequently asked questions **What is investigative tradecraft?** Investigative tradecraft is the disciplined process analysts use to collect information, evaluate sources, test assumptions, and reach defensible conclusions. It combines methodology, experience, and judgment to turn raw information into trusted intelligence. **Why is tradecraft important in intelligence analysis?** Tradecraft is important because it helps analysts separate signal from noise, validate evidence, and avoid unsupported conclusions. In intelligence analysis, strong tradecraft improves accuracy, traceability, and confidence in the final assessment. **What is agentic AI?** Agentic AI refers to AI systems that can plan, execute, and coordinate multi-step tasks with a degree of autonomy. In investigative workflows, agentic AI can help collect, correlate, and organize information while following defined rules and permissions. **How does AI support investigations?** AI supports investigations by accelerating research, surfacing connections across large datasets, summarizing information, and helping analysts prioritize leads. The most effective investigative AI enhances human analysis rather than replacing it. **Why is human judgment important in AI-assisted investigations?** Human judgment is essential because analysts must assess context, challenge assumptions, verify evidence, and decide whether AI-generated findings are credible and fit for purpose. Accountability for investigative conclusions cannot be delegated to AI. **What is the difference between AI and investigative tradecraft?** AI is a technology that can process information and automate parts of an investigation. Investigative tradecraft is the human-led methodology that determines what questions to ask, how evidence should be tested, and whether conclusions are supported. **How does Babel Street apply tradecraft to AI?** Babel Street applies tradecraft to AI by building investigative workflows that emphasize source validation, transparency, repeatability, and human oversight. The goal is to help analysts use AI to scale disciplined intelligence work without losing control of the process. **What can readers expect in Parts 2 and 3 of the Talking Tradecraft series?** [Part 2](https://www.babelstreet.com/blog/talking-tradecraft-part-2-tradecraft-at-scale) explains how Babel Street incorporates tradecraft into Insights Investigator workflows. [Part 3](https://www.babelstreet.com/blog/talking-tradecraft-part-3-how-investigators-think) features practitioner perspectives on the rules and habits that support sound investigative tradecraft.