Can We Trust AI Agents? Understanding Deception, Vulnerabilities & Data Poisoning

Insights from the Bernstein Conference 2026: AI, Data Moats and Market Risk

AI is no longer just a technology story. It is becoming an operating model question. After returning from the Bernstein Strategic Decisions Conference in New York, Montaka’s Lachlan Mackay shared several signals that matter for investors: 1. AI adoption is moving from optional to expected, with companies like MSCI embedding usage into day-to-day work. 2. Proprietary data is becoming more valuable, not less, as AI agents need trusted, real-time inputs. 3. Payment networks such as Visa, Mastercard and Amex are using scale data to strengthen fraud prevention and agent commerce. 4. Private credit is facing a transparency reset, with demand rising for ratings, benchmarks and better asset-level data. 5. Macro risk remains live, even as equity markets remain near record highs. The key question is not “who uses AI?” It is “who owns the data, distribution and trust layer AI needs?” Listen to the latest Montaka Spotlight episode for the full discussion. #GlobalEquities #ArtificialIntelligence #InvestmentInsights #FinancialServices #Montaka Disclaimer: Issued by Montaka Global Pty Ltd ABN 62 604 878 533, AFSL 516942. This information is general in nature and does not take into account your specific needs or circumstances. You should consider your own financial position, objectives and requirements and seek professional financial advice before making any financial decisions.

Where Advantages Lie (And Lie) in the Age of AI

AI agents are being framed as an existential threat to enterprise software, with some investors assuming that software platforms, seat-based models, and switching costs could be eroded almost overnight. In this episode of Montaka’s Spotlight Series, Jessica speaks with Andrew Macken, Chief Investment Officer at Montaka Global Investments, about the central argument from Montaka’s latest whitepaper: Where Advantages Lie in the Age of AI [https://montaka.com/ai-whitepaper/]. They unpack why the AI “doomsday” scenario may be overstated, including the security risks of AI models, the power and compute bottlenecks limiting widespread agent deployment, and the practical difficulty of deploying AI agents safely inside large enterprises. The discussion then turns to where competitive advantages may actually strengthen in an AI world, including scarce compute and electric power, proprietary data flywheels, enterprise distribution, R&D scale, and irreplaceable digital assets. Andrew also explains why some advantaged businesses may have been unfairly sold off as markets price in disruption too broadly. #ArtificialIntelligence #AIInfrastructure #EnterpriseSoftware #CompetitiveAdvantage #InvestmentInsights #GlobalEquities #LongTermInvesting #MontakaGlobal Disclaimer: Issued by Montaka Global Pty Ltd ABN 62 604 878 533, AFSL 516942. This information is general in nature and does not take into account your specific needs or circumstances. You should consider your own financial position, objectives and requirements and seek professional financial advice before making any financial decisions.

Can We Trust AI Agents? Understanding Deception, Vulnerabilities & Data Poisoning

AI agents are becoming more autonomous—but are they trustworthy? In this episode of the Montaka Global Investment Spotlight Series, Chief of Staff Jessica Dharmasiri sits down with CIO Andrew Macken to explore three critical AI safety risks that every business leader needs to understand: 🔍 AI Models Can Deceive – Research shows models deliberately hiding their capabilities to avoid modification 🔍 AI Models Can Be Deceived – Why even extensive guardrails are mathematically insufficient 🔍 Training Data Can Be Poisoned – How just 250 documents can create undetectable backdoors From inbox deletions to system breaches, recent AI agent failures aren't just bugs—they're symptoms of fundamental vulnerabilities. But the solution isn't avoiding AI altogether. Key Takeaways: * Why AI agents demonstrate unexpected situational awareness * The mathematics behind guardrail vulnerabilities * Practical deployment strategies: treating AI agents like unvetted contractors * How to unlock AI value while minimizing enterprise risk Whether you're a CIO evaluating AI deployment or an investment professional tracking the AI revolution, this conversation provides essential insights into the future of autonomous systems. Disclaimer: Issued by Montaka Global Pty Ltd ABN 62 604 878 533, AFSL 516942. This information is general in nature and does not take into account your specific needs or circumstances. You should consider your own financial position, objectives and requirements and seek professional financial advice before making any financial decisions. #ArtificialIntelligence #AI #RiskManagement #BusinessStrategy #TechInvestment

The News is Bad and the Facts are Good

The noise is bad and the facts are good – KKR CEO Scott Nuttall's words from late 2025 perfectly describe many stocks in Montaka's portfolio today. In this Spotlight Series episode, Lachlan Mackay and Tim Le dissect recent financial results from Amazon, Meta, Spotify, REA Group, and DoorDash. Across the board, the results were exceptional and competitive positions continue strengthening – yet stock prices are down significantly. We explore why the market simultaneously fears both AI oversupply AND total disruption – two mutually exclusive scenarios. While short-term prices detach from reality, the team remains focused on fundamental earnings power and sees substantial upside ahead. Disclaimer: Issued by Montaka Global Pty Ltd ABN 62604 878 533, AFSL 516942. This information is general in nature and does not take into account your specific needs or circumstances. You should consider your own financial position, objectives and requirements and seek professional financial advice before making any financial decisions. #Investing #TechStocks #AI #FundamentalAnalysis#ValueInvesting

The SaaS Apocalypse: Why the Market Got It Wrong

We tackle the recent market panic head-on. With software stocks experiencing a dramatic sell-off over AI agent fears, we examine whether the "SaaS apocalypse" narrative holds up to scrutiny. Join Montaka's Director of Research Amit Nath, Senior Analyst Lachlan Mackay, and Portfolio Manager Andrew Macken as they dissect the market's assumption that AI agents will replace traditional enterprise software. Through a candid internal research discussion, we explore: 🔍 Why scale distribution—not code—is the real competitive advantage🔒 The overlooked security risks of unconstrained AI agents⚡ The massive power and compute constraints limiting AI agent deployment📊 Why enterprise customers won't hand over the keys to a single AI provider⏱️ The lengthy implementation timelines that make rapid disruption unlikely Our conclusion? The market has significantly oversold quality enterprise software companies including ServiceNow, Salesforce, Microsoft, MongoDB, and others. This represents a substantial price-to-value divergence that we believe will correct. If you're an investor trying to separate AI hype from reality, this episode is essential listening. #Investing #SoftwareStocks #AI #AIAgents #SaaS #EnterprisesSoftware #StockMarket #ValueInvesting #TechStocks #Salesforce #ServiceNow #Microsoft #MarketAnalysis #InvestmentResearch Disclaimer: Issued by Montaka Global Pty Ltd ABN 62 604 878 533, AFSL 516942. This information is general in nature and does not take into account your specific needs or circumstances. You should consider your own financial position, objectives and requirements and seek professional financial advice before making any financial decisions.