<?xml version="1.0" encoding="UTF-8" ?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
    <channel>
        <atom:link href="https://nanoacademy.ai-thea.com/rss" rel="self" type="application/rss+xml" />
        <title>NanoAcademy by AITHEA</title>
        <link>https://nanoacademy.ai-thea.com</link>
        <description>AITHEA’s NanoAcademy offers a modern, flexible learning experience designed for busy professionals and teams working in risk, trade, finance, and legal roles. Our courses combine expert insights, real-world scenarios, AI and governance, and technology topics in microlearning format to help you build practical skills - one focused lesson at a time.

Whether you're looking to stay ahead of regulatory changes, train your team, or upskill in a specific area, you can pick and choose whatever content you need right now.</description>
        
        <item>
            <guid>https://nanoacademy.ai-thea.com/blog/derisking-compliance</guid>
            <title>De-risking is not compliance. It is risk avoidance dressed as governance.</title>
            <link>https://nanoacademy.ai-thea.com/blog/derisking-compliance</link>
            <description>De-risking isn't compliance — it's avoidance. See why FATF, FinCEN, and the World Bank treat blanket account closures as a regulatory and reputational liability.</description>
            <pubDate>Sun, 20 Sep 2026 22:00:00 +0000</pubDate>                    
            <category>AI</category>
            <category>Compliance</category>
            <category>Tech</category>
            
        </item>
        <item>
            <guid>https://nanoacademy.ai-thea.com/blog/llms-in-case-narrative-generation</guid>
            <title>LLMs in Case Narrative Generation</title>
            <link>https://nanoacademy.ai-thea.com/blog/llms-in-case-narrative-generation</link>
            <description>LLMs can draft AML and KYC case narratives in seconds — and hallucinate transactions that never happened. Where they help, and where they don't.</description>
            <pubDate>Mon, 03 Aug 2026 22:00:00 +0000</pubDate>                    
            <category>Tech</category>
            <category>Compliance</category>
            
        </item>
        <item>
            <guid>https://nanoacademy.ai-thea.com/blog/graph-databases-vs-money-laundering</guid>
            <title>Graph Databases vs. Money Laundering</title>
            <link>https://nanoacademy.ai-thea.com/blog/graph-databases-vs-money-laundering</link>
            <description>Structuring, layering, and funnel accounts hide from transaction rules. See why graph databases catch what rule-based AML monitoring misses.</description>
            <pubDate>Thu, 30 Jul 2026 22:00:00 +0000</pubDate>                    
            <category>Financial Crime</category>
            
        </item>
        <item>
            <guid>https://nanoacademy.ai-thea.com/blog/the-skills-gap-excuse-ends-here</guid>
            <title>The Skills Gap Excuse Ends Here</title>
            <link>https://nanoacademy.ai-thea.com/blog/the-skills-gap-excuse-ends-here</link>
            <description>For a decade, 'we lack the skills' has justified hiring instead of training. Here's why upskilling your compliance team beats hiring a data scientist.</description>
            <pubDate>Thu, 23 Jul 2026 22:00:00 +0000</pubDate>                    
            <category>Compliance</category>
            
        </item>
        <item>
            <guid>https://nanoacademy.ai-thea.com/blog/compliances-precision-recall-edge</guid>
            <title>Compliance's Precision-Recall Edge</title>
            <link>https://nanoacademy.ai-thea.com/blog/compliances-precision-recall-edge</link>
            <description>Why chasing recall without precision quietly burns out your compliance ops team — and what balanced alert design looks like in practice.</description>
            <pubDate>Tue, 21 Jul 2026 22:00:00 +0000</pubDate>                    
            <category>Financial Crime</category>
            <category>Compliance</category>
            <category>AI</category>
            <category>Tech</category>
            
        </item>
        <item>
            <guid>https://nanoacademy.ai-thea.com/blog/what-is-an-roc-curve</guid>
            <title>What is a ROC curve and why does your screening vendor's AUC score hide more than it reveals?</title>
            <link>https://nanoacademy.ai-thea.com/blog/what-is-an-roc-curve</link>
            <description>Area under the ROC curve is a useful summary — at the wrong operating point. How to read it, what it obscures, and the follow-up questions that expose whether a model is genuinely good.</description>
            <pubDate>Sun, 19 Jul 2026 22:00:00 +0000</pubDate>                    
            
        </item>
        <item>
            <guid>https://nanoacademy.ai-thea.com/blog/calibration-vs-accuracy</guid>
            <title>Calibration vs Accuracy: The Model Metric Your Compliance Team Has Never Heard Of</title>
            <link>https://nanoacademy.ai-thea.com/blog/calibration-vs-accuracy</link>
            <description>A model can be 95% accurate and still hand you a confidence score you can't trust. Learn why calibration — not accuracy — is the metric compliance teams need to watch.</description>
            <pubDate>Thu, 16 Jul 2026 22:00:00 +0000</pubDate>                    
            <category>Compliance</category>
            <category>Tech</category>
            
        </item>
        <item>
            <guid>https://nanoacademy.ai-thea.com/blog/youre-managing-a-filing-system-not-financial-crime-risk</guid>
            <title>You're Managing a Filing System, Not Financial Crime Risk</title>
            <link>https://nanoacademy.ai-thea.com/blog/youre-managing-a-filing-system-not-financial-crime-risk</link>
            <description>SARs, KYC records, and case notes weren't built to surface systemic risk — they were built to document individual events. Graph analytics was built for exactly that job.</description>
            <pubDate>Mon, 13 Jul 2026 22:00:00 +0000</pubDate>                    
            <category>Financial Crime</category>
            <category>Tech</category>
            <category>Compliance</category>
            
        </item>
        <item>
            <guid>https://nanoacademy.ai-thea.com/blog/training-a-kyc-risk-scorer</guid>
            <title>Training a KYC Risk Scorer: The Features That Make or Break It</title>
            <link>https://nanoacademy.ai-thea.com/blog/training-a-kyc-risk-scorer</link>
            <description>The model isn't usually the problem — the features are. This piece breaks down how industry, geography, and ownership data get encoded into a KYC risk scorer, and why FATF's 2025 updates and FCA guidance on AI bias make that encoding step a compliance issue, not just a technical one.</description>
            <pubDate>Mon, 06 Jul 2026 22:00:00 +0000</pubDate>                    
            <category>Financial Crime</category>
            <category>AI</category>
            <category>Tech</category>
            <category>Compliance</category>
            
        </item>
        <item>
            <guid>https://nanoacademy.ai-thea.com/blog/when-every-phishing-email-is-personal</guid>
            <title>When Every Phishing Email Is Personal: The New Shape of Fraud</title>
            <link>https://nanoacademy.ai-thea.com/blog/when-every-phishing-email-is-personal</link>
            <description>Generative AI has turned phishing from a mass-produced numbers game into a personalised, industrial-scale operation. This article breaks down how AI-generated lures are feeding account takeover, synthetic identity onboarding, and multimodal BEC — and what FinCEN and FBI guidance says compliance teams should be watching for instead.</description>
            <pubDate>Sun, 05 Jul 2026 22:00:00 +0000</pubDate>                    
            <category>Compliance</category>
            <category>AI</category>
            <category>Tech</category>
            
        </item>
        <item>
            <guid>https://nanoacademy.ai-thea.com/blog/officers-guide-to-ai-vendor-due-diligence</guid>
            <title>Don't Be Sold a Story: The Compliance Officer's Guide to AI Vendor Due Diligence</title>
            <link>https://nanoacademy.ai-thea.com/blog/officers-guide-to-ai-vendor-due-diligence</link>
            <description>Synthetic training data, back-tested results, and single-metric scorecards are standard features of the AI sales process. Here's what they're hiding — and what the EU AI Act, DOJ guidance, and NIST frameworks say your due diligence must cover.</description>
            <pubDate>Sun, 28 Jun 2026 22:00:00 +0000</pubDate>                    
            <category>AI</category>
            <category>Compliance</category>
            <category>Tech</category>
            
        </item>
        <item>
            <guid>https://nanoacademy.ai-thea.com/blog/when-ai-systems-disagree</guid>
            <title>When AI Systems Disagree: The Governance Failure No One Planned For</title>
            <link>https://nanoacademy.ai-thea.com/blog/when-ai-systems-disagree</link>
            <description>Most organisations haven't asked what happens when their AI tools disagree. This post explains why that question is now a compliance risk — and what frameworks like the EU AI Act and NIST RMF say you should do about it.</description>
            <pubDate>Wed, 24 Jun 2026 22:00:00 +0000</pubDate>                    
            <category>AI</category>
            <category>Compliance</category>
            <category>Tech</category>
            
        </item>
        <item>
            <guid>https://nanoacademy.ai-thea.com/blog/can-synthetic-data-solve-your-aml-training-problem</guid>
            <title>Can Synthetic Data Solve Your AML Training Problem?</title>
            <link>https://nanoacademy.ai-thea.com/blog/can-synthetic-data-solve-your-aml-training-problem</link>
            <description>Synthetic data is transforming AML model development — but it has real limits. This article breaks down what it can solve, what it can't replicate, and the questions regulators will ask your team.</description>
            <pubDate>Tue, 16 Jun 2026 22:00:00 +0000</pubDate>                    
            <category>AI</category>
            <category>Financial Crime</category>
            <category>Tech</category>
            
        </item>
        <item>
            <guid>https://nanoacademy.ai-thea.com/blog/the-four-questions</guid>
            <title>The Four Questions Every Compliance Technology Vendor Should Answer Before You Sign</title>
            <link>https://nanoacademy.ai-thea.com/blog/the-four-questions</link>
            <description>In a crowded compliance technology market, knowing the right questions to ask a vendor can be the difference between genuine regulatory protection and expensive exposure. This article breaks down the four non-negotiable questions every organisation should put to a compliance AI vendor before signing: whether their model is truly explainable, where their training data comes from and how it's tracked, how they handle performance degradation when real-world conditions shift, and whether their audit logs are complete, tamper-evident, and under your control. Grounded in EU AI Act obligations and the NIST AI Risk Management Framework, these are the questions regulators are already asking — so you should be asking them first.</description>
            <pubDate>Sun, 14 Jun 2026 22:00:00 +0000</pubDate>                    
            <category>Compliance</category>
            <category>Tech</category>
            
        </item>
        <item>
            <guid>https://nanoacademy.ai-thea.com/blog/financial-crime-compliance-and-ai-governance</guid>
            <title>Financial Crime Compliance and AI Governance Need the Same Person in the Room</title>
            <link>https://nanoacademy.ai-thea.com/blog/financial-crime-compliance-and-ai-governance</link>
            <description>The EU AI Act and AML/CFT frameworks demand many of the same things — explainability, human oversight, audit trails — yet most institutions govern them separately. This article examines what regulators are saying and what integrated governance actually looks like in practice.</description>
            <pubDate>Sun, 31 May 2026 22:00:00 +0000</pubDate>                    
            <category>AI</category>
            <category>Compliance</category>
            <category>Financial Crime</category>
            <category>Tech</category>
            
        </item>
        <item>
            <guid>https://nanoacademy.ai-thea.com/blog/regulatory-technology-compliance-technology</guid>
            <title>Regulatory technology is not the same as compliance technology.</title>
            <link>https://nanoacademy.ai-thea.com/blog/regulatory-technology-compliance-technology</link>
            <description>RegTech and compliance technology are not the same thing — but most institutions treat them as if they are, and the regulatory consequences are well documented. This article breaks down the critical distinction: RegTech automates how you report to regulators; compliance technology manages whether you're actually controlling risk. With global regulatory fines hitting a record $19.3 billion in 2024 and bank penalties surging 522% year-on-year, understanding which tool solves which problem is no longer optional — it's the difference between a functioning compliance programme and an expensive one that still fails.</description>
            <pubDate>Sun, 24 May 2026 22:00:00 +0000</pubDate>                    
            <category>AI</category>
            <category>Compliance</category>
            <category>Tech</category>
            <category>Financial Crime</category>
            
        </item>
        <item>
            <guid>https://nanoacademy.ai-thea.com/blog/the-hallucination-problem-in-compliance-ai</guid>
            <title>The Hallucination Problem in Compliance AI</title>
            <link>https://nanoacademy.ai-thea.com/blog/the-hallucination-problem-in-compliance-ai</link>
            <description>LLMs are being deployed across compliance workflows at speed — but in PEP and sanctions screening, their most dangerous feature isn't what they get wrong. It's that they get it wrong with complete confidence. This article breaks down the hallucination problem in compliance AI, what OFAC, FATF, the FCA, and the EU AI Act actually require of institutions using these tools, and what sound governance looks like before the regulator asks to see it.</description>
            <pubDate>Wed, 20 May 2026 22:00:00 +0000</pubDate>                    
            <category>AI</category>
            <category>Compliance</category>
            <category>Tech</category>
            
        </item>
        <item>
            <guid>https://nanoacademy.ai-thea.com/blog/algorithmic-fairness-in-financial-crime</guid>
            <title>Algorithmic Fairness in Financial Crime</title>
            <link>https://nanoacademy.ai-thea.com/blog/algorithmic-fairness-in-financial-crime</link>
            <description>Algorithmic Fairness in Financial Crime: The Question the Industry Refuses to Ask explores a critical blind spot in AML and fraud detection: AI models trained on historical data can disproportionately flag certain demographics through proxy discrimination, creating both an ethical failure and a growing regulatory exposure. As the EU AI Act (Regulation 2024/1689) brings high-risk AI obligations into force in August 2026 — covering AML compliance and fraud detection — and FATF, the FCA, and the US Treasury all tighten expectations around financial inclusion and de-risking, financial institutions can no longer treat algorithmic fairness as optional. This article examines why the industry has historically looked away, what proxy discrimination looks like in practice, and what compliance and model risk teams must do now: disaggregate false positive rates by demographic proxy, embed fairness testing upstream in model development, and integrate fairness into Model Risk Frameworks. A biased model is an inaccurate model — and in 2026, it is increasingly a non-compliant one.</description>
            <pubDate>Mon, 18 May 2026 22:00:00 +0000</pubDate>                    
            <category>AI</category>
            <category>Compliance</category>
            <category>Financial Crime</category>
            <category>Tech</category>
            
        </item>
        <item>
            <guid>https://nanoacademy.ai-thea.com/blog/agentic-ai</guid>
            <title>Financial services isn't ready for agentic AI</title>
            <link>https://nanoacademy.ai-thea.com/blog/agentic-ai</link>
            <description>New academic research from the University of Edinburgh and UNC Charlotte shows that the safety controls most firms rely on carry no binding enforcement authority at runtime.</description>
            <pubDate>Mon, 11 May 2026 22:00:00 +0000</pubDate>                    
            <category>AI</category>
            <category>Financial Crime</category>
            
        </item>
        <item>
            <guid>https://nanoacademy.ai-thea.com/blog/traditional-vs-ainative</guid>
            <title>Large data publishers vs. AI-native challengers</title>
            <link>https://nanoacademy.ai-thea.com/blog/traditional-vs-ainative</link>
            <description>As generative AI matures, a quiet shift is reshaping the information economy. Established publishers sit on decades of trusted, citable data. New entrants move fast, build natively for AI, and challenge legacy workflows.</description>
            <pubDate>Sun, 10 May 2026 22:00:00 +0000</pubDate>                    
            <category>AI</category>
            <category>Financial Crime</category>
            <category>Tech</category>
            
        </item>
        <item>
            <guid>https://nanoacademy.ai-thea.com/blog/false-positives-transaction-monitoring</guid>
            <title>AML False Positives: What Your Transaction Monitoring Model Is Telling You</title>
            <link>https://nanoacademy.ai-thea.com/blog/false-positives-transaction-monitoring</link>
            <description>Traditional transaction monitoring systems flag up to 95% of alerts as false positives 
— yet most financial institutions treat this as an operational problem rather than a 
model design failure. In this industry perspective, Aithea examines what high false 
positive rates are really telling compliance teams: where thresholds are misset, where 
analyst feedback is lost, and why optimising for alert volume instead of detection 
performance creates dangerous blind spots. With global AML compliance costs exceeding 
$274 billion annually and regulators raising the bar under AMLA and FinCEN's updated 
guidance, the cost of not listening to your model has never been higher.</description>
            <pubDate>Sun, 10 May 2026 22:00:00 +0000</pubDate>                    
            <category>AI</category>
            <category>Compliance</category>
            <category>Financial Crime</category>
            <category>Tech</category>
            
        </item>
        <item>
            <guid>https://nanoacademy.ai-thea.com/blog/name-you-almost-missed</guid>
            <title>Levenshtein, Jaro-Winkler, and the Art of Finding the Name You Almost Missed </title>
            <link>https://nanoacademy.ai-thea.com/blog/name-you-almost-missed</link>
            <description>Not all fuzzy matching algorithms are equal, and in sanctions screening, the difference matters. This article breaks down how Levenshtein and Jaro-Winkler handle real name variations, why your similarity threshold is a compliance decision, and what a robust matching strategy actually looks like. Essential reading for compliance officers, risk teams, and RegTech professionals.</description>
            <pubDate>Wed, 06 May 2026 22:00:00 +0000</pubDate>                    
            <category>AI</category>
            <category>Compliance</category>
            <category>Financial Crime</category>
            <category>Tech</category>
            
        </item>
        <item>
            <guid>https://nanoacademy.ai-thea.com/blog/aml-worst-algortihm</guid>
            <title>Your AML Programme is only as Good as your Worst Algorithm</title>
            <link>https://nanoacademy.ai-thea.com/blog/aml-worst-algortihm</link>
            <description>Your AML programme is only as strong as its weakest algorithm. 

Many compliance teams rely on vendor metrics like accuracy without understanding the underlying model performance. This creates a critical risk gap in AI governance.
This blog explains why the confusion matrix is essential for interpreting AML models, how weak algorithms impact the entire compliance chain, and the key questions teams should be asking vendors. With regulators like the Financial Conduct Authority, Financial Crimes Enforcement Network, and Monetary Authority of Singapore increasing scrutiny on AI, understanding model performance is now a core compliance responsibility.</description>
            <pubDate>Sun, 03 May 2026 22:00:00 +0000</pubDate>                    
            <category>AI</category>
            <category>Compliance</category>
            <category>Financial Crime</category>
            <category>Tech</category>
            
        </item>
        <item>
            <guid>https://nanoacademy.ai-thea.com/blog/eu-omnibus-ai-compliance</guid>
            <title>The EU Digital Omnibus on AI A Double-Edged Sword for Financial Crime Compliance</title>
            <link>https://nanoacademy.ai-thea.com/blog/eu-omnibus-ai-compliance</link>
            <description></description>
            <pubDate>Tue, 28 Apr 2026 22:00:00 +0000</pubDate>                    
            <category>AI</category>
            <category>Financial Crime</category>
            
        </item>
        <item>
            <guid>https://nanoacademy.ai-thea.com/blog/compliance-training-is-ready-for-its-moment</guid>
            <title>Compliance Training Is Ready For Its Moment</title>
            <link>https://nanoacademy.ai-thea.com/blog/compliance-training-is-ready-for-its-moment</link>
            <description>Aithea is rethinking compliance education from the ground up. Learn how NanoAcademy's microlearning approach is helping trade, cargo, and financial crime professionals build knowledge that actually sticks.</description>
            <pubDate>Mon, 27 Apr 2026 22:00:00 +0000</pubDate>                    
            <category>AI</category>
            <category>Compliance</category>
            
        </item>
        <item>
            <guid>https://nanoacademy.ai-thea.com/blog/eu-ai-act</guid>
            <title>The EU AI Act Is a Compliance Framework. For Criminal Networks, It's a Product Spec.</title>
            <link>https://nanoacademy.ai-thea.com/blog/eu-ai-act</link>
            <description>The EU AI Act, AI-enabled fraud, synthetic identity, voice cloning, and money mule networks are converging into a single financial crime threat. Senior compliance and financial crime professionals need to understand the regulatory asymmetry, and how to close it.</description>
            <pubDate>Sun, 26 Apr 2026 22:00:00 +0000</pubDate>                    
            <category>AI</category>
            <category>Compliance</category>
            <category>Financial Crime</category>
            
        </item>
        <item>
            <guid>https://nanoacademy.ai-thea.com/blog/five-ai-algorithms-every-financial-crime-compliance-professional-needs-to-understand</guid>
            <title>Five AI Algorithms Every Financial Crime Compliance Professional Needs to Understand</title>
            <link>https://nanoacademy.ai-thea.com/blog/five-ai-algorithms-every-financial-crime-compliance-professional-needs-to-understand</link>
            <description>AI in financial crime compliance goes far beyond dashboards and detection rates. This practitioner guide explains the five algorithms regulators and risk teams need to understand from gradient boosting in AML transaction monitoring to word embeddings in adverse media screening, and the governance questions every compliance professional should be asking their vendors.</description>
            <pubDate>Wed, 22 Apr 2026 22:00:00 +0000</pubDate>                    
            <category>AI</category>
            <category>Compliance</category>
            <category>Financial Crime</category>
            <category>Tech</category>
            
        </item>
        <item>
            <guid>https://nanoacademy.ai-thea.com/blog/microlearning-in-compliance-learning</guid>
            <title>Why Compliance Training Finally Works: NanoAcademy's Microlearning Approach </title>
            <link>https://nanoacademy.ai-thea.com/blog/microlearning-in-compliance-learning</link>
            <description>Discover how NanoAcademy uses microlearning and gamification to build compliance training courses that employees actually finish and remember. Learn our teaching philosophy. </description>
            <pubDate>Sun, 19 Apr 2026 22:00:00 +0000</pubDate>                    
            <category>Compliance</category>
            <category>AI</category>
            <category>Tech</category>
            
        </item>
        <item>
            <guid>https://nanoacademy.ai-thea.com/blog/explainable-ai-in-compliance</guid>
            <title>Why Explainable AI is the Next Frontier in Trade Compliance </title>
            <link>https://nanoacademy.ai-thea.com/blog/explainable-ai-in-compliance</link>
            <description>AI is transforming trade compliance — but can your team explain the decisions it makes? Explore why Explainable AI (XAI) is becoming a regulatory requirement, what it means for compliance professionals, and how to build the AI literacy your organisation needs to stay ahead.</description>
            <pubDate>Sun, 12 Apr 2026 22:00:00 +0000</pubDate>                    
            
        </item>
        <item>
            <guid>https://nanoacademy.ai-thea.com/blog/compliance-innovation-due-diligence</guid>
            <title>The Compliance Start-up Blind Spot: Why Innovation Gets Lost in Due Diligence</title>
            <link>https://nanoacademy.ai-thea.com/blog/compliance-innovation-due-diligence</link>
            <description>The compliance technology market is trapped in a paradox: organisations urgently need better tools to manage growing regulatory complexity, yet their procurement processes — built for large legacy vendors — systematically filter out the innovative start-ups best placed to help them, through disproportionate certification requirements, stalled sales cycles, and structural bias toward incumbency.
The article argues that fixing this requires buyers to adopt tiered, risk-proportionate procurement frameworks, start-ups to be more strategic about credibility and pilot structures, and the broader ecosystem to create pathways that let innovation actually reach the organisations that need it most.</description>
            <pubDate>Thu, 09 Apr 2026 22:00:00 +0000</pubDate>                    
            <category>Compliance</category>
            <category>AI</category>
            <category>Financial Crime</category>
            <category>Tech</category>
            
        </item>
        <item>
            <guid>https://nanoacademy.ai-thea.com/blog/air-cargo-compliance-frontier</guid>
            <title>Air Cargo’s Next Compliance Frontier: Trade Compliance in an Era of Sanctions, Digitalisation and AI-Driven Risk</title>
            <link>https://nanoacademy.ai-thea.com/blog/air-cargo-compliance-frontier</link>
            <description>A look at how sanctions, digitalisation, and emerging AI-driven risks are reshaping trade compliance in the global air cargo industry.</description>
            <pubDate>Fri, 13 Mar 2026 23:00:00 +0000</pubDate>                    
            
        </item>
    </channel>
</rss>    
        