<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:dl="https://www.delphilens.com/ns/insight"><channel>
  <title>Delphilens Insights</title><link>https://www.delphilens.com/resources/#insights</link>
  <description>Practical insights on taking enterprise AI from strategy to production.</description><language>en-GB</language>
  <atom:link href="https://www.delphilens.com/resources/insights/feed.xml" rel="self" type="application/rss+xml"/>
    <item><title>Trusted Over a Best Friend. Not With the Card.</title><link>https://www.delphilens.com/resources/insights/trusted-not-with-the-card/</link><guid>https://www.delphilens.com/resources/insights/trusted-not-with-the-card/</guid>
      <pubDate>Mon, 16 Nov 2026 00:00:00 +0000</pubDate>
      <category>Financial Services</category><category domain="https://www.delphilens.com/sector">Financial services</category><description>The headline says shoppers are ready to let AI agents buy for them. The same survey says almost nobody will let one pay. That gap sits right where banks and card issuers work.</description>
      <dl:type>Myth vs Reality</dl:type><dl:mins>4</dl:mins><dl:image>https://www.delphilens.com/assets/img/og/trusted-not-with-the-card.jpg</dl:image><dl:takeaway>Accenture’s headline: 74% would trust an AI agent more than their best friend to buy for them.</dl:takeaway><dl:takeaway>The same survey: 32% would let an agent decide what to buy if they pay themselves, and 9% would allow fully autonomous purchases.</dl:takeaway><dl:takeaway>New launches are built for that reality: ChatGPT Wallet asks users to approve every purchase, and Gemini can’t pay from Google Wallet.</dl:takeaway><dl:takeaway>Trust at the payment step depends on what happens when the agent is wrong; Amex now covers errors by registered agents.</dl:takeaway><dl:takeaway>Banks can start now: flag agent-initiated disputes, read network terms, write the customer answer and define the evidence they need.</dl:takeaway></item>
    <item><title>Acts of AI: Price the Disruption Before Your Lawyers Do</title><link>https://www.delphilens.com/resources/insights/acts-of-ai/</link><guid>https://www.delphilens.com/resources/insights/acts-of-ai/</guid>
      <pubDate>Fri, 06 Nov 2026 00:00:00 +0000</pubDate>
      <category>Private Markets</category><category domain="https://www.delphilens.com/sector">Private markets</category><description>Buyout loans have started treating AI like a war or a pandemic. If AI risk is now in the credit agreement, it belongs in your diligence first. Here’s a six-step review a deal team can run in a week.</description>
      <dl:type>Playbook</dl:type><dl:mins>6</dl:mins><dl:image>https://www.delphilens.com/assets/img/og/acts-of-ai.jpg</dl:image><dl:takeaway>Buyout loans now include clauses that treat AI disruption like a pandemic or a war; lenders struck one from a €4.7B deal.</dl:takeaway><dl:takeaway>The 2021 software vintage is under strain: PIK interest, rising loan-to-value and a large maturity wall in 2028–29.</dl:takeaway><dl:takeaway>Seat-priced software doing routine knowledge work is most exposed to AI.</dl:takeaway><dl:takeaway>Run an AI exposure review in six steps: work, seats, customers, management plan, downside case, credit documents.</dl:takeaway><dl:takeaway>Price the risk before signing, so it isn’t negotiated afterwards in the loan documents.</dl:takeaway></item>
    <item><title>Two Weeks to Pack: What You Really Lose When Your AI Coding Tool Is Sold</title><link>https://www.delphilens.com/resources/insights/two-weeks-to-pack/</link><guid>https://www.delphilens.com/resources/insights/two-weeks-to-pack/</guid>
      <pubDate>Thu, 05 Nov 2026 00:00:00 +0000</pubDate>
      <category>Software Delivery</category><category domain="https://www.delphilens.com/sector">Software delivery</category><description>On Wednesday, 100,000 developers learned that the app they use to run AI coding agents had been bought by Meta, and that its cloud would close on 20 November. The code is safe. The way their teams work might not be.</description>
      <dl:type>Field Story</dl:type><dl:mins>4</dl:mins><dl:image>https://www.delphilens.com/assets/img/og/two-weeks-to-pack.jpg</dl:image><dl:takeaway>Meta’s purchase of Conductor gave 100,000 users two weeks before its cloud closes on 20 November.</dl:takeaway><dl:takeaway>Code in Git is safe. Agent instructions, templates, run history and approval records often aren’t.</dl:takeaway><dl:takeaway>Big AI companies are buying developer tools, so more of these deadlines are coming.</dl:takeaway><dl:takeaway>Keep instructions and settings in the repository, export run history, and hold your own model keys.</dl:takeaway><dl:takeaway>Run the two-week test: list what you’d lose that isn’t in Git, then fix it.</dl:takeaway></item>
    <item><title>AI Governance in 2026: What Changed This Year, and What to Do Before 2027</title><link>https://www.delphilens.com/resources/insights/ai-governance-2026/</link><guid>https://www.delphilens.com/resources/insights/ai-governance-2026/</guid>
      <pubDate>Sun, 01 Nov 2026 00:00:00 +0000</pubDate>
      <category>Governance</category><description>The EU delayed its toughest AI rules, India chose guidelines over a new law, and US states kept legislating while Washington deregulated. If that sounds like a reason to relax, it isn’t. Here is what actually changed, and what it means for your organisation.</description>
      <dl:type></dl:type><dl:mins>5</dl:mins><dl:image>https://www.delphilens.com/assets/img/og/ai-governance-2026.jpg</dl:image><dl:takeaway>EU high-risk obligations now start 2 December 2027 (Annex III) and 2 August 2028 (Annex I).</dl:takeaway><dl:takeaway>India governs AI through existing law and guidelines; DPDP duties apply from 13 May 2027.</dl:takeaway><dl:takeaway>US federal policy deregulated while states such as California, Colorado and New York legislated.</dl:takeaway><dl:takeaway>Build an inventory, classify by risk, name owners, evaluate and log.</dl:takeaway><dl:takeaway>Use ISO/IEC 42001 or the NIST AI RMF as a common structure across jurisdictions.</dl:takeaway></item>
    <item><title>Capability That Outlasts the Consultants: Building AI Skills That Stick</title><link>https://www.delphilens.com/resources/insights/ai-capability-that-lasts/</link><guid>https://www.delphilens.com/resources/insights/ai-capability-that-lasts/</guid>
      <pubDate>Wed, 28 Oct 2026 00:00:00 +0000</pubDate>
      <category>Enablement</category><description>Training days are easy to run and easy to forget. Real capability gets built in real work, by the right people at the right depth, and it keeps going after the programme ends.</description>
      <dl:type></dl:type><dl:mins>6</dl:mins><dl:image>https://www.delphilens.com/assets/img/og/ai-capability-that-lasts.jpg</dl:image><dl:takeaway>Match depth to role, from awareness for everyone to leadership for a few.</dl:takeaway><dl:takeaway>Learn on real use cases from the team’s own backlog, and ship something.</dl:takeaway><dl:takeaway>Give champions time, recognition and a direct line to the programme.</dl:takeaway><dl:takeaway>Get leaders using AI visibly, early.</dl:takeaway><dl:takeaway>Measure adoption and outcomes over quarters, not attendance.</dl:takeaway></item>
    <item><title>Plan, Act, Check: Putting AI Agents to Work Without Losing Control</title><link>https://www.delphilens.com/resources/insights/plan-act-check-ai-agents/</link><guid>https://www.delphilens.com/resources/insights/plan-act-check-ai-agents/</guid>
      <pubDate>Sat, 24 Oct 2026 00:00:00 +0000</pubDate>
      <category>Agents</category><description>Agents can take real work off people’s plates. They can also take real actions nobody intended. The answer isn’t to keep them in a box. It’s to design them the way you would any other critical system.</description>
      <dl:type></dl:type><dl:mins>6</dl:mins><dl:image>https://www.delphilens.com/assets/img/og/plan-act-check-ai-agents.jpg</dl:image><dl:takeaway>Use a planner and narrow specialist agents, not one agent with access to everything.</dl:takeaway><dl:takeaway>Put control in the tools: typed, least-privilege, limited, draft first.</dl:takeaway><dl:takeaway>Review every output against rules, policy and evidence before it leaves.</dl:takeaway><dl:takeaway>Make human approvals fast and informative, and reserve them for what matters.</dl:takeaway><dl:takeaway>Build memory, guardrails and observability once, for every agent.</dl:takeaway></item>
    <item><title>The AI Gap in Private Equity: Why Tools Alone Won’t Create Value</title><link>https://www.delphilens.com/resources/insights/ai-enablement-private-equity/</link><guid>https://www.delphilens.com/resources/insights/ai-enablement-private-equity/</guid>
      <pubDate>Tue, 20 Oct 2026 00:00:00 +0000</pubDate>
      <category>Private Markets</category><category domain="https://www.delphilens.com/sector">Private markets</category><description>Private equity firms are buying AI tools faster than they are building the skills to use them. The firms that pull ahead will be the ones that close that gap, in their own teams and across their portfolios.</description>
      <dl:type></dl:type><dl:mins>4</dl:mins><dl:image>https://www.delphilens.com/assets/img/og/ai-enablement-private-equity.jpg</dl:image><dl:takeaway>Most PE firms have AI tools; few have changed how they work.</dl:takeaway><dl:takeaway>Build capability by role (partners, associates, IR) on real deals in a secure environment.</dl:takeaway><dl:takeaway>Treat AI as a portfolio value-creation lever with a consistent readiness review.</dl:takeaway><dl:takeaway>Combine proprietary data with capability; tools alone don’t differentiate.</dl:takeaway><dl:takeaway>Make AI governance part of the diligence and exit story.</dl:takeaway></item>
    <item><title>Why AI Pilots Stall, and Five Decisions That Get Them to Production</title><link>https://www.delphilens.com/resources/insights/why-ai-pilots-stall/</link><guid>https://www.delphilens.com/resources/insights/why-ai-pilots-stall/</guid>
      <pubDate>Wed, 14 Oct 2026 00:00:00 +0000</pubDate>
      <category>Strategy</category><description>Most AI pilots don’t fail because the model is bad. They stall because of decisions nobody wanted to make at the start. Here are the five we now insist on before anyone writes a line of code.</description>
      <dl:type></dl:type><dl:mins>7</dl:mins><dl:image>https://www.delphilens.com/assets/img/og/why-ai-pilots-stall.jpg</dl:image><dl:takeaway>Ask a business leader to own the outcome and the number it should move.</dl:takeaway><dl:takeaway>Agree data access, privacy and residency in plain language before building.</dl:takeaway><dl:takeaway>Judge quality, speed and cost against a graded set of real examples.</dl:takeaway><dl:takeaway>Put AI inside the tools people already use, at the moment they need it.</dl:takeaway><dl:takeaway>Decide who runs it after launch, and build the hand-over into the plan.</dl:takeaway></item>
    <item><title>The Campus AI Gap: Why Institutions Are Behind Their Own Students</title><link>https://www.delphilens.com/resources/insights/ai-in-education-institutions/</link><guid>https://www.delphilens.com/resources/insights/ai-in-education-institutions/</guid>
      <pubDate>Tue, 06 Oct 2026 00:00:00 +0000</pubDate>
      <category>Education</category><category domain="https://www.delphilens.com/sector">Education</category><description>Students adopted AI almost overnight. Most schools, colleges and universities are still deciding what they think about it. That gap is now the biggest risk, and the biggest opportunity, in education.</description>
      <dl:type></dl:type><dl:mins>5</dl:mins><dl:image>https://www.delphilens.com/assets/img/og/ai-in-education-institutions.jpg</dl:image><dl:takeaway>The gap between student adoption and institutional strategy is the real risk.</dl:takeaway><dl:takeaway>Write policies that say what people can do, not only what they can’t.</dl:takeaway><dl:takeaway>Train faculty first; they shape assessment and integrity.</dl:takeaway><dl:takeaway>Redesign assessment deliberately, course by course.</dl:takeaway><dl:takeaway>Provide safe, equitable access grounded in your own content.</dl:takeaway><dl:takeaway>Prepare now for EU high-risk rules (Dec 2027) and India’s DPDP duties (May 2027).</dl:takeaway></item>
</channel></rss>
