{
  "apiVersion": "v1",
  "topic": {
    "id": "ai-cost-and-spend",
    "title": "What does AI cost organizations?",
    "summary": "Spending on models, tokens and tools, and how organizations budget for it.",
    "group": "People and money",
    "aliases": [
      "token spend",
      "ai budget",
      "cost"
    ]
  },
  "passages": [
    {
      "id": "shopify-r-d-majority-internal-ai-spend-allocated-2026",
      "text": "On R&D, the majority of our internal AI spend is allocated here, so you've seen a modest uptick in year-over-year growth.",
      "speaker": "Jeff Hoffmeister",
      "role": "Chief Financial Officer",
      "organization": "Shopify",
      "aiVendor": false,
      "work": {
        "title": "earnings call transcript via Platform Aeronaut (third-party transcript site)",
        "type": "earnings-call",
        "date": "2026-08-05",
        "publisher": "earnings call transcript via Platform Aeronaut (third-party transcript site)",
        "url": "https://transcripts.platformaeronaut.com/transcripts/SHOP-2Q26-transcript"
      },
      "grade": "A",
      "verification": {
        "status": "verified",
        "reviewedOn": "2026-10-31"
      },
      "topics": [
        "ai-cost-and-spend"
      ],
      "functions": [
        "finance"
      ],
      "figures": [],
      "context": null,
      "url": "https://the-ai-record.com/passages/shopify-r-d-majority-internal-ai-spend-allocated-2026"
    },
    {
      "id": "shopify-were-able-hold-gross-margins-relatively-consistent-2026",
      "text": "We were able to hold gross margins at a relatively consistent level quarter over quarter while Sidekiq usage scaled, which reflects some cost efficiencies in support",
      "speaker": "Jeff Hoffmeister",
      "role": "Chief Financial Officer",
      "organization": "Shopify",
      "aiVendor": false,
      "work": {
        "title": "earnings call transcript via Platform Aeronaut (third-party transcript site)",
        "type": "earnings-call",
        "date": "2026-08-05",
        "publisher": "earnings call transcript via Platform Aeronaut (third-party transcript site)",
        "url": "https://transcripts.platformaeronaut.com/transcripts/SHOP-2Q26-transcript"
      },
      "grade": "A",
      "verification": {
        "status": "verified",
        "reviewedOn": "2026-10-31"
      },
      "topics": [
        "ai-cost-and-spend",
        "ai-support-resolution"
      ],
      "functions": [
        "support"
      ],
      "figures": [],
      "context": null,
      "url": "https://the-ai-record.com/passages/shopify-were-able-hold-gross-margins-relatively-consistent-2026"
    },
    {
      "id": "duolingo-inside-business-also-use-ai-more-closer-2026",
      "text": "Inside the business, we also use AI, and that is more closer to the $10 million range internally.",
      "speaker": null,
      "role": null,
      "organization": "Duolingo",
      "aiVendor": false,
      "work": {
        "title": "earnings call transcript via Webull (third-party transcript site)",
        "type": "earnings-call",
        "date": "2026-08",
        "publisher": "earnings call transcript via Webull (third-party transcript site)",
        "url": "https://www.webull.com/news/15392596720296960"
      },
      "grade": "A",
      "verification": {
        "status": "verified",
        "reviewedOn": "2026-10-31"
      },
      "topics": [
        "ai-cost-and-spend"
      ],
      "functions": [
        "company-wide"
      ],
      "figures": [],
      "context": null,
      "url": "https://the-ai-record.com/passages/duolingo-inside-business-also-use-ai-more-closer-2026"
    },
    {
      "id": "klarna-decreased-spending-external-marketing-suppliers-such-translation-2026",
      "text": "We decreased spending on external marketing suppliers, such as translation, production, CRM and social agencies, by $7 million in the first quarter of 2024 (as compared to the first quarter of 2023) and we attribute 37% of this decrease to the use of AI",
      "speaker": null,
      "role": null,
      "organization": "Klarna",
      "aiVendor": false,
      "work": {
        "title": "SEC filing (the company's own regulatory report)",
        "type": "filing",
        "date": "2026-02-26",
        "publisher": "SEC filing (the company's own regulatory report)",
        "url": "https://www.sec.gov/Archives/edgar/data/2003292/000200329226000007/klar-20251231.htm"
      },
      "grade": "A",
      "verification": {
        "status": "verified",
        "reviewedOn": "2026-10-31"
      },
      "topics": [
        "ai-cost-and-spend",
        "sales-and-marketing"
      ],
      "functions": [
        "sales-marketing"
      ],
      "figures": [],
      "context": null,
      "url": "https://the-ai-record.com/passages/klarna-decreased-spending-external-marketing-suppliers-such-translation-2026"
    },
    {
      "id": "xebia-planning-agent-runs-your-most-powerful-2026",
      "text": "The planning agent is responsible for the actual thinking, so it should run on your most powerful model. I initially used Fable for this purpose when it was available, as it excels in UI/UX design. When Fable was no longer accessible, I switched to Opus. Since the planner writes the specifications, the implementation can be handled by a more cost-effective and faster model like Sonnet.",
      "speaker": "Noah Blauensteiner",
      "role": "Author, Xebia articles",
      "organization": "Xebia",
      "aiVendor": true,
      "work": {
        "title": "Shining Some Light in Dark Agentic Software Factories",
        "type": "post",
        "date": "2026-08-12",
        "publisher": "Xebia articles",
        "locator": "What is an Agentic Software Factory?",
        "url": "https://xebia.com/articles/shining-some-light-in-dark-agentic-software-factories/"
      },
      "grade": "B",
      "verification": {
        "status": "verified",
        "reviewedOn": "2026-10-31"
      },
      "topics": [
        "coding-agents",
        "ai-cost-and-spend"
      ],
      "functions": [
        "engineering"
      ],
      "figures": [],
      "context": null,
      "url": "https://the-ai-record.com/passages/xebia-planning-agent-runs-your-most-powerful-2026"
    },
    {
      "id": "xebia-running-several-agents-parallel-burns-tokens-2026",
      "text": "Running several agents in parallel burns tokens quickly, so set budgets and keep an eye on spending from day one.",
      "speaker": "Noah Blauensteiner",
      "role": "Author, Xebia articles",
      "organization": "Xebia",
      "aiVendor": true,
      "work": {
        "title": "Shining Some Light in Dark Agentic Software Factories",
        "type": "post",
        "date": "2026-08-12",
        "publisher": "Xebia articles",
        "locator": "What is an Agentic Software Factory?",
        "url": "https://xebia.com/articles/shining-some-light-in-dark-agentic-software-factories/"
      },
      "grade": "B",
      "verification": {
        "status": "verified",
        "reviewedOn": "2026-10-31"
      },
      "topics": [
        "ai-cost-and-spend"
      ],
      "functions": [
        "engineering"
      ],
      "figures": [],
      "context": null,
      "url": "https://the-ai-record.com/passages/xebia-running-several-agents-parallel-burns-tokens-2026"
    },
    {
      "id": "boundary-ab-test-tool-calls-billions-of-tokens-programming-language-2026",
      "text": "You could figure out which ones took less tool calls, … which one made less errors, which one produced the correct outcome, and deterministically know what's going on. The point is you can start building data-driven systems without ever writing a single line of code. … in order to build a programming language, it wouldn't have taken eight people. It wouldn't have taken less than two years. It would have taken hundreds and thousands and tens of thousands of man hours, and then you would still have a broken system. And today, we can just spend billions of tokens and make it work, and we can make it stable.",
      "speaker": "Vaibhav Gupta",
      "role": "Co-founder and Chief Executive",
      "organization": "Boundary (BoundaryML)",
      "aiVendor": true,
      "work": {
        "title": "Fighting Slop with Slop",
        "type": "talk",
        "date": "2026-07-31",
        "publisher": "AI Engineer (YouTube)",
        "locator": "6:00",
        "url": "https://www.youtube.com/watch?v=AMiyLItEtLA"
      },
      "grade": "B",
      "verification": {
        "status": "verified",
        "reviewedOn": "2026-11-01"
      },
      "topics": [
        "measuring-impact",
        "developer-productivity",
        "ai-cost-and-spend"
      ],
      "functions": [
        "engineering"
      ],
      "figures": [],
      "context": "Gupta is comparing what building a programming language would have taken before with what it takes the team now.",
      "url": "https://the-ai-record.com/passages/boundary-ab-test-tool-calls-billions-of-tokens-programming-language-2026"
    },
    {
      "id": "reddit-flag-agent-seven-for-seven-green-ci-1-26-per-pull-request-2026",
      "text": "Before I trusted this workflow, I back tested the scoring against months of clean up history. Then I ran it live. Seven for seven PRs with green CI. Now it runs daily on my laptop and hands work back for me to review. … $1.26 a pull request. A backlog of around 520 flags a year runs for under $700. If you had engineers working on this by hand, at least $26,000 that you're looking at.",
      "speaker": "Andrew Orobator",
      "role": "Senior Software Engineer",
      "organization": "Reddit",
      "aiVendor": false,
      "work": {
        "title": "Scale the Judgment, Not the Model",
        "type": "talk",
        "date": "2026-06-30",
        "publisher": "AI Engineer (YouTube)",
        "locator": "11:59",
        "url": "https://www.youtube.com/watch?v=6MudaeKdBSk"
      },
      "grade": "B",
      "verification": {
        "status": "verified",
        "reviewedOn": "2026-11-01"
      },
      "topics": [
        "coding-agents",
        "ai-cost-and-spend",
        "measuring-impact"
      ],
      "functions": [
        "engineering"
      ],
      "figures": [
        {
          "metric": "flag-cleanup-prs-green-ci",
          "value": 7,
          "unit": "count",
          "period": "the live run after back testing",
          "definition": "pull requests from the speaker's stale-feature-flag agent that passed CI, out of 7 (\"Seven for seven\")",
          "level": "human-approves"
        },
        {
          "metric": "flag-cleanup-cost-per-pr",
          "value": 1.26,
          "unit": "USD",
          "period": "per pull request",
          "definition": "cost per pull request of the speaker's stale-feature-flag agent, as he states it",
          "level": "human-approves"
        },
        {
          "metric": "flag-cleanup-yearly-cost",
          "value": 700,
          "unit": "USD",
          "period": "per year, backlog of around 520 flags",
          "definition": "upper bound ('under $700') the speaker gives for running his agent on about 520 flags a year",
          "level": "human-approves"
        },
        {
          "metric": "flag-cleanup-manual-cost",
          "value": 26000,
          "unit": "USD",
          "period": "per year, same backlog",
          "definition": "the speaker's estimate ('at least $26,000') of the same work done by engineers by hand",
          "level": "n/a"
        }
      ],
      "context": "The speaker is a Reddit Android engineer. In the talk he describes this agent as running in his local workflow and on his laptop, and says some of his other practices are from his personal side projects. The $26,000 is his own estimate of manual cost, and the PR count is a small sample.",
      "url": "https://the-ai-record.com/passages/reddit-flag-agent-seven-for-seven-green-ci-1-26-per-pull-request-2026"
    },
    {
      "id": "uber-we-saw-against-like-very-naive-2026",
      "text": "we saw that against like a very naive implementation, our costs were down by 60%, and our quality and our accuracy was up by around 70%.",
      "speaker": "Ameya Ketkar",
      "role": "Speaker, AI Engineer World's Fair 2026 (job title not stated)",
      "organization": "Uber",
      "aiVendor": false,
      "work": {
        "title": "Building uReview, Uber's Multi-Agent Code Review Engine (talk, AI Engineer World's Fair 2026)",
        "type": "talk",
        "date": "2026-08-28",
        "publisher": "AI Engineer (YouTube)",
        "locator": "10:50",
        "url": "https://www.youtube.com/watch?v=EL123UNokkI"
      },
      "grade": "B",
      "verification": {
        "status": "reported"
      },
      "topics": [
        "code-review",
        "ai-cost-and-spend",
        "measuring-impact"
      ],
      "functions": [
        "engineering"
      ],
      "figures": [
        {
          "metric": "ai-review-cost-change",
          "value": -60,
          "unit": "percent",
          "period": "against a naive first implementation",
          "definition": "cost of running uReview, compared with a very naive implementation, as Ketkar describes it",
          "level": "n/a"
        },
        {
          "metric": "ai-review-quality-change",
          "value": 70,
          "unit": "percent",
          "period": "against a naive first implementation",
          "definition": "quality and accuracy of uReview, compared with a very naive implementation, as Ketkar describes it; the talk does not say how this is measured",
          "level": "n/a"
        }
      ],
      "context": "Ameya Ketkar is summarizing the effect of the observability and evaluation work on uReview, Uber's code-review system.",
      "url": "https://the-ai-record.com/passages/uber-we-saw-against-like-very-naive-2026"
    },
    {
      "id": "cloudflare-security-agent-processes-over-7-billion-tokens-2026",
      "text": "we have a security agent that processes over 7 billion tokens per day on Kimi. That would cost an estimated $2.4M per year on a mid-tier proprietary model. But on Workers AI, it’s 77% cheaper.",
      "speaker": null,
      "role": null,
      "organization": "Cloudflare",
      "aiVendor": true,
      "work": {
        "title": "The AI engineering stack we built internally, on the platform we ship",
        "type": "post",
        "date": "2026-04-20",
        "publisher": "Cloudflare blog",
        "locator": "Act 1, \"How we increasingly leverage Workers AI\"",
        "url": "https://blog.cloudflare.com/internal-ai-engineering-stack/"
      },
      "grade": "B",
      "verification": {
        "status": "reported"
      },
      "topics": [
        "ai-cost-and-spend",
        "security"
      ],
      "functions": [
        "engineering"
      ],
      "figures": [
        {
          "metric": "tokens-per-day",
          "value": 7000000000,
          "unit": "count",
          "period": "per day, security agent on Kimi K2.5",
          "definition": "Tokens processed by an internal security agent (stated as over 7 billion)",
          "level": "n/a"
        },
        {
          "metric": "cost-saving",
          "value": 0.77,
          "unit": "share",
          "period": "versus an estimated $2.4M per year on a mid-tier proprietary model",
          "definition": "Stated saving from running the security agent on Workers AI, Cloudflare's own product",
          "level": "n/a"
        }
      ],
      "context": "Cloudflare sells Workers AI, the product it describes running this agent on.",
      "url": "https://the-ai-record.com/passages/cloudflare-security-agent-processes-over-7-billion-tokens-2026"
    },
    {
      "id": "klarna-supported-ai-enabled-productivity-gains-continued-cost-2026-2",
      "text": "supported by AI-enabled productivity gains and continued cost discipline.",
      "speaker": null,
      "role": null,
      "organization": "Klarna",
      "aiVendor": false,
      "work": {
        "title": "SEC filing (the company's own regulatory report), furnished on Form 6-K",
        "type": "filing",
        "date": "2026-08-18",
        "publisher": "SEC filing (the company's own regulatory report), furnished on Form 6-K",
        "url": "https://www.sec.gov/Archives/edgar/data/2003292/000162828026057573/q22026earningsrelease.htm"
      },
      "grade": "A",
      "verification": {
        "status": "verified",
        "reviewedOn": "2026-10-31"
      },
      "topics": [
        "ai-cost-and-spend"
      ],
      "functions": [
        "company-wide"
      ],
      "figures": [],
      "context": null,
      "url": "https://the-ai-record.com/passages/klarna-supported-ai-enabled-productivity-gains-continued-cost-2026-2"
    },
    {
      "id": "airbnb-customer-support-related-cost-per-booking-declined-2026",
      "text": "our customer support related cost per booking declined approximately 16% year-over-year",
      "speaker": null,
      "role": null,
      "organization": "Airbnb",
      "aiVendor": false,
      "work": {
        "title": "SEC filing (8-K exhibit: the company's own shareholder letter)",
        "type": "filing",
        "date": "2026-08-06",
        "publisher": "SEC filing (8-K exhibit: the company's own shareholder letter)",
        "url": "https://www.sec.gov/Archives/edgar/data/0001559720/000119312526337928/d70413dex991.htm"
      },
      "grade": "A",
      "verification": {
        "status": "verified",
        "reviewedOn": "2026-10-31"
      },
      "topics": [
        "ai-cost-and-spend",
        "ai-support-resolution"
      ],
      "functions": [
        "support"
      ],
      "figures": [],
      "context": null,
      "url": "https://the-ai-record.com/passages/airbnb-customer-support-related-cost-per-booking-declined-2026"
    },
    {
      "id": "airbnb-its-not-about-how-many-tokens-you-2026",
      "text": "it's not about how many tokens you use, it's about the throughput of your product and the quality of your product designs and how much you're shipping.",
      "speaker": "Brian Chesky",
      "role": "Co-founder and Chief Executive Officer",
      "organization": "Airbnb",
      "aiVendor": false,
      "work": {
        "title": "earnings call transcript via Platform Aeronaut (third-party transcript site)",
        "type": "earnings-call",
        "date": "2026-08-06",
        "publisher": "earnings call transcript via Platform Aeronaut (third-party transcript site)",
        "url": "https://transcripts.platformaeronaut.com/transcripts/ABNB-2Q26-transcript"
      },
      "grade": "A",
      "verification": {
        "status": "verified",
        "reviewedOn": "2026-10-31"
      },
      "topics": [
        "developer-productivity",
        "ai-cost-and-spend",
        "measuring-impact"
      ],
      "functions": [
        "company-wide"
      ],
      "figures": [],
      "context": null,
      "url": "https://the-ai-record.com/passages/airbnb-its-not-about-how-many-tokens-you-2026"
    },
    {
      "id": "airbnb-obviously-does-assume-material-increase-terms-ai-2026",
      "text": "it obviously does assume a material increase in terms of the AI spend over the course of the year",
      "speaker": "Ellie Mertz",
      "role": "Chief Financial Officer",
      "organization": "Airbnb",
      "aiVendor": false,
      "work": {
        "title": "earnings call transcript via Platform Aeronaut (third-party transcript site)",
        "type": "earnings-call",
        "date": "2026-08-06",
        "publisher": "earnings call transcript via Platform Aeronaut (third-party transcript site)",
        "url": "https://transcripts.platformaeronaut.com/transcripts/ABNB-2Q26-transcript"
      },
      "grade": "A",
      "verification": {
        "status": "verified",
        "reviewedOn": "2026-10-31"
      },
      "topics": [
        "ai-cost-and-spend"
      ],
      "functions": [
        "finance"
      ],
      "figures": [],
      "context": null,
      "url": "https://the-ai-record.com/passages/airbnb-obviously-does-assume-material-increase-terms-ai-2026"
    },
    {
      "id": "uber-customer-support-clear-area-where-should-able-2026",
      "text": "Customer support is a clear area where we should be able to up the quality of our customer support interactions as well as reduce the cost of our effort there.",
      "speaker": "Balaji Krishnamurthy",
      "role": "Chief Financial Officer",
      "organization": "Uber",
      "aiVendor": false,
      "work": {
        "title": "earnings call transcript via Platform Aeronaut (third-party transcript site)",
        "type": "earnings-call",
        "date": "2026-08-05",
        "publisher": "earnings call transcript via Platform Aeronaut (third-party transcript site)",
        "url": "https://transcripts.platformaeronaut.com/transcripts/UBER-2Q26-transcript"
      },
      "grade": "A",
      "verification": {
        "status": "verified",
        "reviewedOn": "2026-10-31"
      },
      "topics": [
        "ai-cost-and-spend",
        "ai-support-resolution"
      ],
      "functions": [
        "support"
      ],
      "figures": [],
      "context": null,
      "url": "https://the-ai-record.com/passages/uber-customer-support-clear-area-where-should-able-2026"
    },
    {
      "id": "klarna-supported-ai-enabled-productivity-gains-continued-cost-2026",
      "text": "This is supported by AI-enabled productivity gains and continued cost discipline across the business.",
      "speaker": null,
      "role": null,
      "organization": "Klarna",
      "aiVendor": false,
      "work": {
        "title": "SEC filing (the company's own regulatory report), furnished on Form 6-K",
        "type": "filing",
        "date": "2026-05-14",
        "publisher": "SEC filing (the company's own regulatory report), furnished on Form 6-K",
        "url": "https://www.sec.gov/Archives/edgar/data/2003292/000162828026034877/exhibitno992q126earnings.htm"
      },
      "grade": "A",
      "verification": {
        "status": "verified",
        "reviewedOn": "2026-10-31"
      },
      "topics": [
        "developer-productivity",
        "ai-cost-and-spend"
      ],
      "functions": [
        "company-wide"
      ],
      "figures": [],
      "context": null,
      "url": "https://the-ai-record.com/passages/klarna-supported-ai-enabled-productivity-gains-continued-cost-2026"
    },
    {
      "id": "ibm-ai-agent-helped-contribute-40-reduction-hr-2025",
      "text": "The AI agent helped contribute to a 40% reduction in the HR team’s operational costs over the past four years.",
      "speaker": null,
      "role": null,
      "organization": "IBM",
      "aiVendor": true,
      "work": {
        "title": "IBM (own site)",
        "type": "post",
        "date": "2025",
        "publisher": "IBM (own site)",
        "url": "https://www.ibm.com/case-studies/ibm-askhr"
      },
      "grade": "B",
      "verification": {
        "status": "verified",
        "reviewedOn": "2026-10-31"
      },
      "topics": [
        "ai-cost-and-spend",
        "headcount-and-hiring"
      ],
      "functions": [
        "people"
      ],
      "figures": [],
      "context": null,
      "url": "https://the-ai-record.com/passages/ibm-ai-agent-helped-contribute-40-reduction-hr-2025"
    }
  ],
  "notice": [
    "Every passage is reproduced word for word from the cited source.",
    "Check the verification status: \"reported\" means not yet confirmed by an independent second reader.",
    "Organizations flagged aiVendor sell AI tooling and have a reason to publish flattering numbers.",
    "An empty result means the archive holds nothing on it, not that nobody has said anything."
  ]
}