Skip to content
Slow Variables
Ask

Adaptation · Adopters

FDA-authorised AI-enabled devices, growth

Year-on-year growth in the cumulative count of AI-enabled medical devices with an FDA final decision, from the FDA's published list. Derived metric fda_ai_devices_yoy.

Reason through this

The reading

Authorisations are mostly radiology CAD tools; the list says what clears review, not what is used at the bedside.

Every readingeach dot links to its record
68.5% · 2019-01-31 · measured66.3% · 2019-02-28 · measured64.6% · 2019-03-31 · measured64.1% · 2019-04-30 · measured60.6% · 2019-05-31 · measured57.4% · 2019-06-30 · measured61.3% · 2019-07-31 · measured63.1% · 2019-08-31 · measured64.8% · 2019-09-30 · measured62.3% · 2019-10-31 · measured58.6% · 2019-11-30 · measured52.6% · 2019-12-31 · measured53.5% · 2020-01-31 · measured57.0% · 2020-02-29 · measured60.7% · 2020-03-31 · measured58.0% · 2020-04-30 · measured55.4% · 2020-05-31 · measured55.8% · 2020-06-30 · measured53.6% · 2020-07-31 · measured52.8% · 2020-08-31 · measured54.4% · 2020-09-30 · measured55.9% · 2020-10-31 · measured52.7% · 2020-11-30 · measured49.1% · 2020-12-31 · measured49.6% · 2021-01-31 · measured43.9% · 2021-03-31 · measured45.7% · 2021-04-30 · measured44.9% · 2021-05-31 · measured43.3% · 2021-06-30 · measured41.4% · 2021-07-31 · measured40.5% · 2021-08-31 · measured39.0% · 2021-09-30 · measured37.1% · 2021-10-31 · measured36.0% · 2021-11-30 · measured37.3% · 2021-12-31 · measured37.4% · 2022-01-31 · measured37.5% · 2022-02-28 · measured35.0% · 2022-03-31 · measured35.7% · 2022-04-30 · measured37.8% · 2022-05-31 · measured38.9% · 2022-06-30 · measured39.3% · 2022-07-31 · measured39.1% · 2022-08-31 · measured37.3% · 2022-09-30 · measured37.3% · 2022-10-31 · measured36.2% · 2022-11-30 · measured34.1% · 2022-12-31 · measured31.7% · 2023-01-31 · measured32.5% · 2023-02-28 · measured34.4% · 2023-03-31 · measured34.3% · 2023-04-30 · measured33.3% · 2023-05-31 · measured30.8% · 2023-06-30 · measured30.1% · 2023-07-31 · measured29.0% · 2023-08-31 · measured30.3% · 2023-09-30 · measured32.3% · 2023-10-31 · measured33.4% · 2023-11-30 · measured35.5% · 2023-12-31 · measured37.3% · 2024-01-31 · measured35.7% · 2024-02-29 · measured33.8% · 2024-03-31 · measured32.7% · 2024-04-30 · measured32.5% · 2024-05-31 · measured32.8% · 2024-06-30 · measured31.0% · 2024-07-31 · measured32.6% · 2024-08-31 · measured30.8% · 2024-09-30 · measured29.1% · 2024-10-31 · measured28.5% · 2024-11-30 · measured27.2% · 2024-12-31 · measured27.5% · 2025-01-31 · measured29.0% · 2025-03-31 · measured28.4% · 2025-04-30 · measured29.9% · 2025-05-31 · measured30.7% · 2025-06-30 · measured31.5% · 2025-07-31 · measured30.4% · 2025-08-31 · measured31.1% · 2025-09-30 · measured30.6% · 2025-10-31 · measured30.1% · 2025-11-30 · measured30.5% · 2025-12-31 · measured29.5% · 2026-01-31 · measured28.2% · 2026-02-28 · measured29.2% · 2026-03-31 · measured28.6% · 2026-04-30 · measured27.3% · 2026-05-31 · measured26.7% · 2026-06-30 · measured
a readingnormal rangefast range
measured

88 readings drawn from the derived metric fda_ai_devices_yoy; 102 observations in the evidence.

Derived (fda_ai_devices_yoy) from: FDA AI-enabled medical device list. FDA, Artificial Intelligence-Enabled Medical Devices list.

The numbers
as ofreadingvaluerecord
2026-06-30—26.7%measured
2026-05-31—27.3%measured
2026-04-30—28.6%measured
2026-03-31—29.2%measured
2026-02-28—28.2%measured
2026-01-31—29.5%measured
2025-12-31—30.5%measured
2025-11-30—30.1%measured
2025-10-31—30.6%measured
2025-09-30—31.1%measured
2025-08-31—30.4%measured
2025-07-31—31.5%measured
2025-06-30—30.7%measured
2025-05-31—29.9%measured
2025-04-30—28.4%measured
2025-03-31—29.0%measured
2025-01-31—27.5%measured
2024-12-31—27.2%measured
2024-11-30—28.5%measured
2024-10-31—29.1%measured
2024-09-30—30.8%measured
2024-08-31—32.6%measured
2024-07-31—31.0%measured
2024-06-30—32.8%measured
2024-05-31—32.5%measured
2024-04-30—32.7%measured
2024-03-31—33.8%measured
2024-02-29—35.7%measured
2024-01-31—37.3%measured
2023-12-31—35.5%measured
2023-11-30—33.4%measured
2023-10-31—32.3%measured
2023-09-30—30.3%measured
2023-08-31—29.0%measured
2023-07-31—30.1%measured
2023-06-30—30.8%measured
2023-05-31—33.3%measured
2023-04-30—34.3%measured
2023-03-31—34.4%measured
2023-02-28—32.5%measured
2023-01-31—31.7%measured
2022-12-31—34.1%measured
2022-11-30—36.2%measured
2022-10-31—37.3%measured
2022-09-30—37.3%measured
2022-08-31—39.1%measured
2022-07-31—39.3%measured
2022-06-30—38.9%measured
2022-05-31—37.8%measured
2022-04-30—35.7%measured
2022-03-31—35.0%measured
2022-02-28—37.5%measured
2022-01-31—37.4%measured
2021-12-31—37.3%measured
2021-11-30—36.0%measured
2021-10-31—37.1%measured
2021-09-30—39.0%measured
2021-08-31—40.5%measured
2021-07-31—41.4%measured
2021-06-30—43.3%measured
2021-05-31—44.9%measured
2021-04-30—45.7%measured
2021-03-31—43.9%measured
2021-01-31—49.6%measured
2020-12-31—49.1%measured
2020-11-30—52.7%measured
2020-10-31—55.9%measured
2020-09-30—54.4%measured
2020-08-31—52.8%measured
2020-07-31—53.6%measured
2020-06-30—55.8%measured
2020-05-31—55.4%measured
2020-04-30—58.0%measured
2020-03-31—60.7%measured
2020-02-29—57.0%measured
2020-01-31—53.5%measured
2019-12-31—52.6%measured
2019-11-30—58.6%measured
2019-10-31—62.3%measured
2019-09-30—64.8%measured
2019-08-31—63.1%measured
2019-07-31—61.3%measured
2019-06-30—57.4%measured
2019-05-31—60.6%measured
2019-04-30—64.1%measured
2019-03-31—64.6%measured
2019-02-28—66.3%measured
2019-01-31—68.5%measured

Why it matters

The clearest regulated-domain speed limit with an official monthly series; the paper predicts slow diffusion in high-consequence tasks and this is the count of what gets through.

What cuts against it

The list is updated in batches and reclassified; a cleared device is not a deployed one; LLM-based devices are a handful at most.

How it is measured
  • derived metric fda_ai_devices_yoy (formula in the semantic layer)
  • source FDA AI-enabled medical device list · default tier 4 · Public domain (U.S. government work)
  • proxy policy · cadence quarterly · valve adoption to adaptation

Normal = the cumulative count growing under 50% a year (the 2019-2024 pace); fast = doubling or faster, which is what an ungated regulated domain would show. Between is emerging.

Applied to metric:fda_ai_devices_yoy.

Derived rows (88)
as ofdimsvalueinputs
2026-06-3026.7%sources12
2026-05-3127.3%sources12
2026-04-3028.6%sources12
2026-03-3129.2%sources12
2026-02-2828.2%sources12
2026-01-3129.5%sources12
2025-12-3130.5%sources12
2025-11-3030.1%sources12
2025-10-3130.6%sources12
2025-09-3031.1%sources12
2025-08-3130.4%sources12
2025-07-3131.5%sources12
2025-06-3030.7%sources12
2025-05-3129.9%sources12
2025-04-3028.4%sources12
2025-03-3129.0%sources12
2025-01-3127.5%sources12
2024-12-3127.2%sources12
2024-11-3028.5%sources12
2024-10-3129.1%sources12
2024-09-3030.8%sources12
2024-08-3132.6%sources12
2024-07-3131.0%sources12
2024-06-3032.8%sources12
2024-05-3132.5%sources12
2024-04-3032.7%sources12
2024-03-3133.8%sources12
2024-02-2935.7%sources12
2024-01-3137.3%sources12
2023-12-3135.5%sources12
2023-11-3033.4%sources12
2023-10-3132.3%sources12
2023-09-3030.3%sources12
2023-08-3129.0%sources12
2023-07-3130.1%sources12
2023-06-3030.8%sources12
2023-05-3133.3%sources12
2023-04-3034.3%sources12
2023-03-3134.4%sources12
2023-02-2832.5%sources12
2023-01-3131.7%sources12
2022-12-3134.1%sources12
2022-11-3036.2%sources12
2022-10-3137.3%sources12
2022-09-3037.3%sources12
2022-08-3139.1%sources12
2022-07-3139.3%sources12
2022-06-3038.9%sources12
2022-05-3137.8%sources12
2022-04-3035.7%sources12
2022-03-3135.0%sources12
2022-02-2837.5%sources12
2022-01-3137.4%sources12
2021-12-3137.3%sources12
2021-11-3036.0%sources12
2021-10-3137.1%sources12
2021-09-3039.0%sources12
2021-08-3140.5%sources12
2021-07-3141.4%sources12
2021-06-3043.3%sources12
2021-05-3144.9%sources12
2021-04-3045.7%sources12
2021-03-3143.9%sources12
2021-01-3149.6%sources12
2020-12-3149.1%sources12
2020-11-3052.7%sources12
2020-10-3155.9%sources12
2020-09-3054.4%sources12
2020-08-3152.8%sources12
2020-07-3153.6%sources12
2020-06-3055.8%sources12
2020-05-3155.4%sources12
2020-04-3058.0%sources12
2020-03-3160.7%sources12
2020-02-2957.0%sources12
2020-01-3153.5%sources12
2019-12-3152.6%sources12
2019-11-3058.6%sources12
2019-10-3162.3%sources12
2019-09-3064.8%sources12
2019-08-3163.1%sources12
2019-07-3161.3%sources12
2019-06-3057.4%sources12
2019-05-3160.6%sources12
2019-04-3064.1%sources12
2019-03-3164.6%sources12
2019-02-2866.3%sources12
2019-01-3168.5%sources12
Latest readings
  • 2026-06-30 · 26.7%as of 2026-06-30(2 obs)
  • 2026-05-31 · 27.3%as of 2026-05-31(2 obs)
  • 2026-04-30 · 28.6%as of 2026-04-30(2 obs)
  • 2026-03-31 · 29.2%as of 2026-03-31(2 obs)
  • 2026-02-28 · 28.2%as of 2026-02-28(2 obs)
  • 2026-01-31 · 29.5%as of 2026-01-31(2 obs)
Why this status, and its history

Evaluator: 26.69 (FDA AI-enabled medical device list, 2026-06-30) is inside the consistent band (hi=50). Auto-reason; band rationale: Normal = the cumulative count growing under 50% a year (the 2019-2024 pace); fast = doubling or faster, which is what an ungated regulated domain would show. Between is emerging.

  1. 2026-09-10unmeasured to consistent with normalconf — → 70 · the evaluator (automatic)

    Evaluator: 26.69 (FDA AI-enabled medical device list, 2026-06-30) is inside the consistent band (hi=50). Auto-reason; band rationale: Normal = the cumulative count growing under 50% a year (the 2019-2024 pace); fast = doubling or faster, which is what an ungated regulated domain would show. Between is emerging.

    sources12
Why this confidence

Rests on 102 observations from FDA AI-enabled medical device list; best evidence tier 4, grade A; current.

Confidence is independent of status: 90–95 multiple strong independent sources; 70–89 good evidence, some ambiguity; 50–69 mixed or hard to operationalise; 0–49 limited or vague evidence.

Predictions that lean on this