The AI Identity Risk Your Board Can’t See | Okta SVP & GM Dan Mountstephen

Machine Customers Are Here. Your Business Was Built for Humans | Katja Forbes

Machine Customers Are Here. Your Business Was Built for Humans | Katja Forbes, Author of The Machine Customers. Subscribe: If your board is talking AI strategy but nobody has raised machine customers, this episode is the briefing. Your Next Customer Won't Be Human: 5 Machine Buyer Types Hitting Australian Commerce. Adobe reported a 4,700% year-on-year increase in AI agent traffic to retail websites. That is not a forecast. That already happened. Your commerce channels, your checkout, your fraud stack, your entire customer journey were built for humans. There are five distinct types of non-human buyer already trying to transact in the Australian economy, and most mid-market businesses are blocking revenue they do not even know is walking in. Katja Forbes is a executive director at Standard Chartered's corporate and investment bank, AFR 100 Women of Influence, and author of The Machine Customers. She has been mapping the taxonomy of machine buyers since before most boards acknowledged the category existed. If your board's AI strategy does not account for the fact that your next customer might not be human, it has a gap. What you'll learn: * The five machine customer types already in market: co-buyer, delegated agent, autonomous buyer, multi-agent network, intermediary broker, and why each one needs a different receptor * Why Salesforce just removed the human UI layer entirely and rearchitected its platform for AI agents as the primary actors * The shift from KYC to KYA: what agent verification looks like and why American Express is now covering losses from registered agent errors * How Kyriba's agentic AI is already executing autonomous FX transactions inside a $208 trillion cross-border payments flow * What went wrong when an AI agent was given $100K and told to open a retail store Key stats: * 4,700% YoY increase in AI agent traffic to retail websites (Adobe). * $30 trillion machine customer economy forecast by 2030 (Gartner). * $208 trillion in cross-border payment flows, $625 billion in bank revenue. Banks that build headless receptors for machine customers will amplify that number. 48-hour action: Identify which of the five machine customer types is most likely to arrive at your business first. Check whether your current channels can serve it. Discoverability gets you into the consideration set. Operational trust is the minimum. Values-based differentiation is what makes the agent choose you over the competitor. "When everyone's discoverable, everyone becomes interchangeable." - Katja Forbes Timestamps: * 00:00 - The machine customer economy hits $30 trillion * 01:00 - Adobe: 4,700% AI agent traffic increase * 05:00 - The five types of machine customer * 09:00 - Amazon Rufus, Woolworths Olive, and the neutrality problem * 13:00 - Salesforce goes headless * 15:00 - Receptors: why each type needs a different customer journey * 19:00 - From KYC to KYA: agent verification and privacy * 22:00 - Ramon's own machine commerce misfire * 26:00 - The $208T cross-border FX opportunity * 31:00 - Luna's $100K autonomous business experiment * 33:00 - Anthropic: Opus vs Haiku trading outcomes * 36:00 - Your 48-hour action * 38:00 - Close www.appliedaiaustralia.com.au [https://www.appliedaiaustralia.com.au]linkedin.com/in/ramonrod [https://www.linkedin.com/company/108631445/admin/dashboard/] Guest:Katja Forbes, Author of The Machine Customers | AFR 100 Women of Influence | Former ED, Standard Charteredlinkedin.com/in/katjaforbes [linkedin.com/in/katjaforbes ]Book: amazon.com/dp/1923630008 [amazon.com/dp/1923630008]Machine Customer Canvas (free): thecxevolutionist.ai/resources/machine-customer-canvas About Applied AI Australia:We help Australian companies between $100M and $1B turn AI into revenue, margin, time back, and better operating discipline. One podcast and one newsletter each week, built so you can brief a board in under an hour.

The AI Identity Risk Your Board Can’t See | Okta SVP & GM Dan Mountstephen

The AI Risk Your Board Can’t See: Identity, Fraud and AI Agents | Dan Mountstephen, Okta Subscribe - know a director who still thinks cyber is an IT problem? Forward this. CBA self-reported $1B in suspected AI-generated fraud this year and committed $900M to fight back. For every human logging into your systems, there are 82 machine identities. Service accounts, API tokens, AI agents. Most have no owner. Most have more access than the people they serve. When one of them goes rogue, the board's looking at you. Dan Mountstephen runs Asia Pacific for Okta, a $3B identity-security business. 20 years in enterprise tech, the last five deep in identity. The biggest risk in your business isn't the people you hire. It's the accounts you forgot about. What you'll learn: * The 82:1 ratio: how machine identities outnumber human ones, and why nobody's watching * Why AI isn't breaking new systems. It's scaling the weaknesses you already have. * How to govern an AI agent the same way you'd govern a contractor * The kill-switch: what to do when an agent goes rogue * Why directors are personally on the hook if due diligence can't be proven Key stats: * 82:1 machine to human identity ratio. * 86% of staff using unsanctioned AI. * 22% of incidents start with stolen credentials. 48-hour action: Get visibility of every identity across your organisation, human and machine. You can't secure what you can't see. "AI isn't breaking systems. It's scaling the weaknesses you already have." - Dan Mountstephen Timestamps: 1. 0:00 - 82 machines per human 2. 1:00 - CBA: $1B fraud, $900M response 3. 3:00 - AI scales weaknesses, doesn’t create them 4. 5:00 - 212% YoY cybercrime growth 5. 5:10 - 86% using unsanctioned AI 6. 6:00 - What good cyber looks like 7. 7:00 - Identity Security Posture Management 8. 8:00 - Directors personally on the hook 9. 10:00 - Static API keys = standing privilege 10. 12:00 - Kill chain: identity to damage 11. 13:00 - Agents as contractors: the ownership model 12. 16:00 - Dell case study 13. 18:00 - Where to start: visibility first 14. 19:00 - 48-hour action 15. 20:15 - Blast radius and kill switch 16. 22:00 - Close www.appliedaiaustralia.com.au linkedin.com/in/ramonrod Guest: Dan Mountstephen, SVP & GM APAC at Oktalinkedin.com/in/danmountstephen | okta.com About Applied AI Australia: We help Australian companies between $100m and $1b turn AI into revenue, margin, time back, and better operating discipline. One podcast and one newsletter each week, built so you can brief a board in under an hour.

60% Fewer People, 90% Faster: The Vacancy Strategy (Diego Mogollon, VentureCrowd)

Fewer People, More Output: The Vacancy Strategy (Diego Mogollon, VentureCrowd) Subscribe: Know a leader still hiring the same org chart? Forward this. You can drive AI transformation in compliance heavy industry! This episode proves it! Diego Mogollon wears two hats at VentureCrowd: CMO and CTO. The regulated FinTech platform has raised $440 million. * Over the last two years, * Tech team shrank 60%. * Development time dropped 90%. * Marketing team went from six people to one, and they're handling more volume than ever before. He didn't fire anyone. His team kept getting poached because they were good. Each time someone left, instead of opening a job req, Diego mapped every task that person owned and asked three questions: 1. what can an agent handle 2. who can upskill into this, 3. what process can we redesign He calls it the Vacancy Strategy, it's changed everything about how VentureCrowd builds. What you'll learn: * Why backfilling roles the traditional way is the biggest missed opportunity in AI adoption * The Customer Zero strategy: build AI agents for your team before your customers * How a crew of specialised AI agents replaced a marketing team of six in a regulated FinTech * Agent KPIs most companies aren't tracking: resolution rates, escalation rates, latency, and context poisoning * Why giving agents more data makes them worse, not better * The shift from human in the loop to human at the helm * How to run a 24-hour agent challenge with your team this week Key frameworks: * Vacancy Strategy: Every time someone leaves, don't backfill. Map. Reassess. Redesign. * Customer Zero: Build AI for your internal team first. Solve your problems, then your customers'. * Agent KPIs: Resolution rate, escalation rate, latency, tokens per agent. * Context Poisoning: An agent should know what it needs to know. You don't dump the company wiki on a new hire's first day. Key stats: * Tech team 60% smaller over two years * Development time down 90% (months to weeks, weeks to days) * Marketing team: 6 people to 1 * $440M raised through VentureCrowd's digital platform * Bug debugging time cut 90%+ with a single agent 48-hour action: Next time a role opens on your team, hold the job req for 48 hours. Map every task that person did. Identify one that an agent could handle. You'll find it. Guest: Diego Mogollon, CMO/CTO at VentureCrowd LinkedIn: linkedin.com/in/diegomogollon Company: venturecrowd.com.au Host: Ramon Rodriguez LinkedIn: linkedin.com/in/ramonrodriguez [linkedin.com/in/ramonrodriguez] Website: www.appliedaiaustralia.com.au

AI Ready in 2026 - Part 2 (Anthony Mittelmark)

$28M Spent on AI. One Question Tells You If It's Working. Subscribe - Know a CFO still measuring AI by project count? Forward this. Episode Summary: 90% of Australian companies have invested in AI. Only 40% are seeing results. The problem isn't the technology. In Part 1, we showed you which type of AI leader you are: Believer, Driver, or Builder. In Part 2, we reveal why that's only half the picture. There are two tracks of AI maturity and most organisations only measure one. Anthony Mittelmark calls it the Two-Track Trap: your AI usage might be advancing, but if your organisation isn't built to carry it, the investment leaks value. This episode gives you diagnostic questions most leadership teams can't answer, a real case study of how one multinational reframed AI as a valuation play, and a 48-hour action you can run this week without budget or approval. What you will learn: * Why treating AI projects like IT projects is the #1 reason programs stall * The four levels of AI usage maturity (Tool User to AI-First) and where most companies are stuck * The second maturity track nobody measures: organisational readiness * Why tactical AI integration can make the rest of the business worse * How a multinational used upstream data IP to change their valuation narrative * A 48-hour brutal audit you can run with your leadership team this week Key Frameworks: Two-Track Trap Track A: AI Usage Maturity (Tool User > Process Runner > Commercial Operator > AI-First) Track B: Organisational Maturity (leadership articulation, business resistance, architecture to roadmap) Most companies measure Track A and ignore Track B. That's where investment leaks. Portfolio Mindset: Every AI project must deliver commercial value, build reusable capability, and advance organisational AI fluency. Valuation Reframe: Not "what does your org look like in 2030" but "how will it be valued in 2030?" Key Stats: * 12% of Australian leaders say GenAI is transforming their business (vs 25% globally) * 90% invested in AI, only 40% seeing results (Frontier, March 2026) * Financial services moving to AI-first: cost-to-serve gap makes it impossible to compete without it 48 Hour Challenge: Ramon: 30 mins. One page. Can leadership articulate why you're investing in AI (commercial reasons, not efficiencies)? Is the business resisting - where and who? What does next-gen CX look like? Does tech map AI architecture to a roadmap with cost-benefit? For every AI initiative, does the org support it or is tech running ahead? If you can't answer with confidence, there's a gap. Anthony: List every AI tool, subscription, pilot, and POC. Portfolio strategy or just a collection of stuff? CEO to CFO: are we aligned on AI investment and what's the return? "The board doesn't need another AI update. They need a name to it next quarter." Timestamps * 0:00 - Intro: the two-track blind spot * 1:00 -Why AI programs stall: IT projects vs portfolio mindset * 5:00 - The four levels of AI usage maturity * 8:30 - The Two-Track Trap: organisational readiness * 11:30 - What happens when leadership can't articulate why * 14:30 - Level 3 usage, Level 1 organisation * 17:30 - The $500M CEO building a business within her business * 19:00 - Upstream data IP and valuation narrative * 21:30 - Does this project advance your AI thesis? * 23:00 - The brutal audit * 25:30 - Close: AI maturity runs on two tracks Part 1 (Believer, Driver, Builder): Here [https://open.spotify.com/episode/6a8VWPYDUyl6XXPUcSfak9?si=0d58b44b26134bc2] Applied AI Australia: * https://appliedaiaustralia.com.au * https://www.linkedin.com/in/ramonrod * Podcast + Executive Newsletter [https://substack.com/@appliedaiaustralia] Guest: Anthony Mittelmark, 20+ years AI strategy across ASX 100 companies and high-growth ventures.. * LinkedIn: linkedin.com/in/anthonymittelmark [https://linkedin.com/in/anthonymittelmark]

AI Ready in 2026: Know Your Type (Part 1)

AI Ready in 2026: Know Your Type (Part 1) Applied AI Australia In Conversation With: Anthony Mittelmark. Technology executive with 20+ years delivering enterprise-scale AI strategy and digital transformation across ASX 100 companies and high-growth ventures. Duration: ~28 minutes EPISODE SUMMARY Every Australian executive falls into one of three AI leadership categories. The problem: most think they're a different type to what they actually are. Ramon Rodriguez and Anthony Mittelmark unpack the Believer, the Driver, and the Builder using real data from KPMG, Deloitte, and MIT. Part 1 is diagnosis. Part 2 covers what to do about it. THE 3 TYPES Type 1: The Believer Well read, forwarding articles to leadership, but hasn't committed to a first commercial move. The risk of being wrong feels bigger than the risk of waiting. That's a false trade-off. Every quarter at Level 1 widens the gap. Type 2: The Driver Committed, shipping results, capital deployed, reorgs done. But running from fear of being caught flat-footed. The tension between AI's 9-18 month value curve and 6-month ASX reporting cycles pushes Drivers toward highly visible, low-impact tools over structural redesign. Type 3: The Builder Budget, pilots, task force, enthusiasm. Missing the commercial thread. Has AI champions but no AI owners. The champion raises their hand. The owner answers to the board when it fails. KEY DATA * 63% of Australian C-suite rank AI as chief concern for 2026 (KPMG) * 12% of Australian leaders report AI is actively transforming their business, vs 25% globally (Deloitte) * 28% of Australian organisations have moved 40%+ of AI pilots to production (Deloitte) * 95% of enterprise gen AI pilots fail to deliver measurable impact (MIT) THE EXECUTIVE LANGUAGE DECODER * "Exploring AI opportunities" = We haven't committed to anything * "Piloting in one department" = We don't know if it works yet * "Monitoring the landscape" = We're waiting for someone else to go first * "Developing our AI strategy" = We don't have one * "AI-augmented workflows" = We added a chatbot * "Responsible AI approach" = We're using compliance as a brake YOUR 48-HOUR FIRST MOVE Believers: Map your category against a 20% margin compression scenario. One page. This week. Have the conversation. Drivers: Ask your leadership team: where are we still making decisions the same way we did three years ago? Find it before a competitor does. Builders: Name one AI initiative running right now and put a strict dollar figure on it. Savings or revenue. If you can't do this by Friday, pause the build. TIMESTAMPS 00:00 Introduction and episode framing 01:07 The Australian AI readiness gap in numbers 03:11 The 3 types of AI leader: overview 03:37 Type 1: The Believer 09:26 Believer self-diagnosis questions 11:00 Type 2: The Driver 12:42 Tactical AI vs AI-first 16:09 Type 3: The Builder 22:00 The Executive Language Decoder 23:56 Traditional frameworks vs probabilistic technology 26:08 Your 48-hour first move 27:20 Closing and Part 2 preview LINKS Website: www.appliedaiaustralia.com.au Executive Newsletter: https://substack.com/@appliedaiaustralia Ramon Rodriguez: https://www.linkedin.com/in/ramonrod Anthony Mittelmark: https://www.linkedin.com/in/anthonymittelmark/ SUBSCRIBE TO APPLIED AI AUSTRALIA The #1 source for Applied AI strategy for the Australian C-Suite. Frameworks you can apply in 48 hours.. ABOUT APPLIED AI AUSTRALIA Applied AI Australia helps execs cut through the noise and deliver practical outcomes: growth, margins, and time. We translate AI into executive action every Monday and Friday.