What will be the largest potentially_affected value for any NHTSA recall campaign received from June 1 through August 11, 2026?

• My most likely prediction for the largest potentially_affected value is 841,999.84 units, with a 50% chance of falling between 495,499.91 and 1,683,499.67 units.
• The forecast is heavily influenced by Ford’s recent recall dominance; the company recalled 19.6 million vehicles in the year ending March 2026, including single campaigns exceeding 4.3 million units.
• Modern recalls are increasingly driven by software issues and Over-the-Air (OTA) updates, which allow manufacturers to address millions of vehicles simultaneously in a single filing.
• Historical data shows that while recalls exceeding 1 million units are common (roughly 10-12 per year), campaigns exceeding 4 million units are rare ‘tail’ events, typically occurring only once every 1.5 to 2 years.
• The 72-day window (June 1 to August 11) is likely to capture at least one major ‘platform-wide’ recall from a top-tier manufacturer like Ford, Tesla, or Toyota.
• Regulatory pressure remains high, with Ford operating under a 2024 NHTSA consent order that mandates more aggressive safety reporting and oversight.

Key figures

Historical context

• Ford (Feb 2026): Campaign 26V104 affected 4,380,609 vehicles (F-150, Maverick, Expedition) due to software glitches in lighting and trailer brakes.
• Tesla (Dec 2023 - Jan 2024): Two separate software-based recalls affected 2,031,220 vehicles (Autopilot) and over 2,000,000 vehicles (warning lights), respectively.
• General Motors (June 2014): Announced a single recall of 8.45 million cars as part of its ignition switch crisis; this remains a benchmark for hardware-based mega-recalls.
• Toyota (2009-2010): Sequential recalls for floor mats and accelerator pedals affected 4.2 million and 2.3 million vehicles, respectively.
• Annual Volume Trends: NHTSA reported 31.3M affected units in 2025 and 35.0M in 2024. Q1 2026 alone saw 11.6 million units, indicating a very active recall environment entering mid-2026.

Tailwinds

• Software-Defined Vehicles: The shift to common software architectures across entire fleets (like Tesla’s FSD or Ford’s integrated trailer modules) means a single bug can affect 2-4 million vehicles at once.
• Regulatory Scrutiny: The 2024 NHTSA consent order for Ford and ongoing investigations into Tesla’s Autopilot and Ford’s BlueCruise increase the likelihood of massive “voluntary” filings to avoid penalties.
• Aging Hardware Populations: Millions of vehicles with aging Takata or ARC airbag inflators remain on the road; a new defect finding in these populations could immediately trigger a 2M+ unit filing.

Headwinds

• Improved Quality Controls: Ford and GM have both reported billion-dollar targets for warranty cost reductions in 2026, which could lead to fewer high-volume defect reports.
• Divergence of Components: If manufacturers split large-scale problems into multiple smaller, model-specific recall campaigns rather than one ‘global’ filing, the maximum ‘potentially_affected’ value per filing will be lower.
• Short Window Duration: The 72-day window is only ~20% of the year; there is a statistical possibility that no major multi-million unit defect is reported specifically during these ten weeks.

Detailed reasoning

My analysis suggests that the most likely value for the largest single recall in the specified window is approximately 841,999 units, with a 50% chance of the value falling between 495,500 and 1,683,500 units. This forecast is grounded in three distinct operational regimes for the NHTSA dataset during a 72-day period.

First, I considered the ‘Base Rate’ regime. Historically, the U.S. sees roughly 1,000 recalls per year. In any 72-day window (roughly 20% of a year), approximately 200 recall reports are filed. Statistical analysis of the maximum value in such a sample size suggests that at least one recall will typically cross the 500,000-unit threshold, often involving a major manufacturer like Ford, GM, or Toyota.

Second, I analyzed ‘High-Volume Platform’ recalls. Recent data from 2024–2026 shows a significant uptick in recalls affecting between 1 million and 2.2 million units. These are often driven by software defects (OTA updates) or component failures on ‘global platforms’ that share parts across multiple models. Ford, for instance, had several such recalls in the spring of 2026 (1.4 million F-150s in April and 1.7 million vehicles for cameras in March). Tesla also frequently enters this range with its Autopilot and warning light software updates (2.0M to 2.2M units). Given that Ford recalled 19.6 million vehicles in the 12 months preceding March 2026, there is a strong probability that a million-plus unit campaign will hit the NHTSA desk during the forecast window.

Third, I accounted for ‘Mega-Campaign’ tail risks. Occasionally, a single report covers an entire production generation or a massive safety crisis (like the Takata airbag or GM ignition switch events). Ford’s February 2026 recall of 4.38 million vehicles is a prime recent example. SEC filings indicate that both Ford and NHTSA are still monitoring massive populations of airbag inflators (up to 3.5 million units) that could trigger a single massive filing. While these ‘mega’ events are rarer, the current heightened regulatory pressure—including a 2024 consent order for Ford—increases the probability of such a filing.

My weighting favors the 500k to 1.7M range (Component 1 and 2), as a 72-day window is a relatively short period, and while million-unit recalls are frequent, 4-million-unit recalls usually only happen once every 12-18 months. The positive skew in my forecast reflects the reality that recall populations are bounded by zero but have a long right tail towards several million units.

Key uncertainties

• Manufacturer Reporting Strategies: Manufacturers may choose to bundle several related issues into one large filing or split them into multiple smaller reports, directly affecting the “largest value” metric.
• OTA vs. Physical Recalls: While software recalls (OTA) scale more easily to millions of units, NHTSA’s classification of these as “recalls” is the primary driver of high numbers.
• Specific Large-Scale Investigations: Current investigations into Tesla’s HW3 hardware and Ford’s hands-free systems have unknown resolution dates; a filing in late June would drastically shift the outcome.

Conclusion

• Decision-makers should expect the largest single recall between June and August 2026 to fall between 495,500 and 1.68 million units, with Ford and Tesla remaining the most likely sources of these high-volume campaigns.
• While a “mega-recall” exceeding 4 million units is possible given the 2026 climate, it is statistically less likely than a standard million-unit platform issue during this specific 72-day window.

Question: what is the largest “potentially_affected” value among DOT/NHTSA recall rows with report_received_date between 2026-06-01 and 2026-08-11? The teams agree this is an order-statistic taken from the DOT Recalls Data (Socrata) table, not a news count, and that comparable June–Aug windows historically produce maxima typically in the mid-six-figure to low-seven-figure range with occasional multi-million outliers. Historical analogues (Honda, FCA, Ford) and recent 2026 Ford recalls show seven-figure and even 4.38M campaigns are possible, so forecasters place the median near ~1M with most mass between ~0.5M–2M. Several analysts flag live Tesla FSD investigations covering ~2.9–3.2M vehicles as the clearest route to a multi-million-row filing in the window, producing a meaningful 2.5–4M tail. The main counterweight—agreed by teams—is timing and conversion risk: large investigations often close without a recall and many big 2026 campaigns already occurred before June 1. Overall probabilities: small chance (<~5–10%) of a max below ~50k, substantial mass in 0.5–2M, meaningful tail around 2–4M, and a small open upper-tail probability above 5M.

Forecast and Reasoning

Forecast: Probability distribution:

• 1.00% chance of value below 90000.0
• 5.00% chance of value below 180000.0
• 10.00% chance of value below 400000.0
• 20.00% chance of value below 700000.0
• 30.00% chance of value below 1000000.0
• 40.00% chance of value below 1250000.0
• 50.00% chance of value below 1550000.0
• 60.00% chance of value below 1900000.0
• 70.00% chance of value below 2400000.0
• 80.00% chance of value below 3200000.0
• 90.00% chance of value below 4200000.0
• 95.00% chance of value below 5250000.0
• 99.00% chance of value below 6200000.0

Forecast Reasoning

We are forecasting the maximum (single-campaign) potentially_affected among all NHTSA recall rows whose report_received_date falls between 2026-06-01 and 2026-08-11 (inclusive). Because this is a maximum over a fairly busy ~72-day window, the distribution is heavy-tailed: most campaigns are small, but the period’s maximum is often driven by whether one “mega-recall” (software/OTA or high-volume platform issue) lands in-window.

Across the four analyst forecasts, the central tendency for the maximum is consistently in the ~1.1–2.1 million range (medians: 1.1M, 2.1M, 1.7M, 1.2M). All analysts also anchor the upper tail on recent multi-million campaigns (notably a documented ~4.38M Ford campaign earlier in 2026), implying a meaningful chance the window’s maximum reaches 3–4+ million, and a non-trivial chance of exceeding the 5,000,000 in-range cap (handled via the open-upper tail).

I weight Forecasters 1/3/4 slightly more than Forecaster 2 because Forecaster 2’s median and upper tail look a bit aggressive for a 72-day window and they show an internal inconsistency on the p99 value vs the platform’s maximum usable past-upper value. Netting this out, I set a median near ~1.55M, with p80 around ~3.2M, p90 around ~4.2M, and I represent about ~5% probability of exceeding 5,000,000 by placing p95 in the open-above region.

A common amateur mistake here is to anchor on typical recall sizes (tens of thousands) rather than the maximum over a multi-month window, which systematically underestimates the outcome. Another common mistake is to over-anchor on a single headline mega-recall and implicitly assume 5M+ campaigns are frequent. My forecast improves calibration by (1) explicitly treating the target as an extreme value over many recalls, (2) using the team’s consistent empirical anchors that recent maxima can reach ~4.4M, and (3) expressing open-upper tail risk without letting it dominate the median.

Forecast rationale (numeric):

— Iteration 1 — Across the forecasts, the main reasoning is that this 72-day window is long enough to likely contain at least one sizable NHTSA recall, and recall sizes are extremely skewed, so the maximum in a short period is often driven by a single large campaign rather than the average recall.

Key factors used

• Historical base rates: Forecasts lean heavily on past NHTSA recall distributions, especially how often million-plus campaigns appear in a 2–3 month span.
• Recent 2026 activity: Strong early-2026 recall volume is used as evidence that another large campaign is plausible within the June 1–August 11 window.
• Heavy-tailed recall sizes: There is broad agreement that recall sizes are highly uneven, with rare mega-recalls possible even if most campaigns are much smaller.
• Window length: A 10.4-week period is treated as long enough to make a large maximum likely, but still short enough that the exact peak is highly uncertain.
• Data/definition uncertainty: Some downside risk is attributed to quiet periods, campaign splitting, or parsing/exclusion effects that could suppress the observed maximum.

Areas of consensus

• The central expectation is in the million-vehicle range, not tens of thousands.
• The distribution should have a wide lower tail and an even wider upper tail.
• Most probability mass is believed to remain below 5 million, while still allowing a small chance of a much larger outlier.

Main differences

• The forecasts differ mainly in how large they think the typical maximum will be:
  • one centers around 1.3M
  • another around 1.5M
  • another around 2.05M
• They also vary in how strongly they weight the chance of a very large campaign, with some emphasizing a more moderate “large recall” scenario and others giving more room for a rare mega-recall.

Overall synthesis

The shared view is that the largest recall in this period will likely be a large, possibly multi-million-vehicle campaign, with recent recall activity and historical skew making a million-scale maximum the most likely outcome. At the same time, the forecasts keep substantial uncertainty because a single exceptional OEM or supplier recall could push the maximum far above the central estimate.

— Iteration 2 — Across the forecasts, the main reasoning pattern is to anchor the estimate in the historical distribution of the largest NHTSA recall received within a roughly 72-day window and then adjust for the chance that the period includes one unusually large campaign.

Shared conclusions

• The most likely outcome is that the largest recall in the window will be around 1–2 million potentially affected vehicles.
• The forecasts treat the distribution as strongly right-skewed/heavy-tailed, meaning most periods have a moderate maximum, but a small number can produce very large outliers.
• A key driver of the upper tail is the possibility of a major OEM recall, especially software/OTA-related recalls or broad component defects that can affect very large fleets.
• The window is short enough that the largest recall is expected to be large but usually not extreme.

Points of disagreement

• The central estimate varies somewhat, from roughly 1.15 million to 1.8 million, reflecting different interpretations of recent recall frequency and the likelihood of a major event.
• The forecasts differ most on the extreme upper tail:
  • One view allows for very large but still bounded rare outcomes.
  • Another assigns meaningful probability to 10M+ or even 25M+ outlier events, citing historical precedents like major safety crises or fleet-wide software issues.
• The lower tail is treated more conservatively by some than others, but all agree that very small maxima are possible only in unusually quiet periods.

Bottom line The collective view is that the period will most likely contain at least one substantial recall near the low millions, with uncertainty driven less by ordinary recall activity and more by whether a rare, very large campaign occurs.

— Iteration 3 — Across the forecasts, the reasoning is dominated by a heavy-tailed view of NHTSA recall sizes: most recall campaigns in a 72-day window are expected to be moderate, but the largest campaign is driven by the chance of one unusually large event.

Main factors used

• Recent precedent: The Q1 2026 Ford recall of about 4.38 million units is treated as an important anchor for what can happen in a short window.
• Typical window behavior: A 2–3 month period is expected to include many recalls, but the maximum is usually thought to land in the 1–3 million range.
• Rare mega-recall risk: The upper tail allows for a systemic defect, software issue, or multi-model/supplier problem that could create a much larger campaign.
• Guidance constraints: The forecasts explicitly incorporate the idea that the value is likely above 50,000 and likely below 5 million, while still leaving room for rare outliers.

Areas of consensus

• The distribution is strongly skewed, with a long right tail.
• A large recall in the low millions is the central expectation.
• Most probability mass sits below 5 million, even though much larger values remain possible.
• The lower end is kept realistic by allowing for a quieter period with only small campaigns.

Main differences in emphasis

• One forecast is more conservative, centering around 1.45 million and treating very large outcomes as possible but not central.
• Others place somewhat more weight on mega-recall scenarios, with tails extending into the 10–30 million+ range, though still as rare events.
• The upper-tail uncertainty varies, but all forecasts agree that the maximum could be far above ordinary recall sizes if a major defect emerges.

Unified takeaway

The collective reasoning suggests that the largest potentially affected value in this window is most likely in the low millions, with the most plausible maximum around 1–3 million and a meaningful chance of something nearer 4–5 million, while still leaving a small probability of a much larger, exceptional recall.

Synthesis of Forecasting Rationales

Historical Base Rates and Seasonality Forecasters ground their analysis in the historical base rates of the 72-day resolution window (representing roughly one-fifth of the calendar year). Historically, a window of this duration is highly likely to capture at least one “mega-recall.” Furthermore, summer months frequently experience a clustering of large defect announcements driven by mid-year regulatory deadlines, end-of-quarter internal corporate audits, and routine Early Warning Reporting (EWR) requirements.

Structural Industry Shifts A major consensus across the rationales highlights a structural shift that is inflating the size of modern recalls: the proliferation of Over-The-Air (OTA) software updates. Because modern vehicles rely on shared electronic control units and common software architectures, a single coding anomaly can instantly compromise an automaker’s entire model lineup. The NHTSA classifies these mandatory software patches as standard safety recalls, significantly raising the ceiling for single-campaign vehicle counts.

Active Regulatory Investigations (The Inside View) Forecasters heavily weight several active NHTSA Engineering Analyses (EAs) that are maturing and could force massive regulatory actions during the summer window. The most prominent catalysts include:

• Tesla’s Full Self-Driving (FSD) Probe: A recently escalated investigation covering a massive fleet that could easily result in a preemptive or mandated fleet-wide OTA recall.
• Honda Defect Probes: Multiple maturing investigations—specifically regarding Auto Idle Stop and Automatic Emergency Braking (AEB) systems—affecting millions of vehicles. These investigations are closely aligning with standard regulatory resolution timelines.
• Ford Investigations: Active probes into systems like BlueCruise or engine downshifting that also cover extensive vehicle populations.

Tail Risks and Administrative Quirks To account for extreme outliers, rationales point to the persistent threat of supplier-level equipment failures (such as widespread airbag or fuel pump defects) which can occasionally trigger unprecedented, multi-million-unit recalls. Conversely, forecasters note that administrative lags in the DOT public portal might artificially truncate the observable timeframe, as defects filed in the final days of the window may not be published in time to be captured by the system.

SUMMARY

Question: What will be the largest potentially_affected value for any NHTSA recall campaign received from June 1 through August 11, 2026? Final Prediction: Probability distribution:

• 10.00% chance of value below 345000.0
• 20.00% chance of value below 551666.666667
• 40.00% chance of value below 915000.0
• 60.00% chance of value below 1420000.0
• 80.00% chance of value below 2666666.666667
• 90.00% chance of value below 3966666.666667

Total Cost: extra_metadata_in_explanation is disabled Time Spent: extra_metadata_in_explanation is disabled LLMs: extra_metadata_in_explanation is disabled Bot Name: extra_metadata_in_explanation is disabled

Report 1 Summary

Forecasts

Forecaster 1: Probability distribution:

• 10.00% chance of value below 355000.0
• 20.00% chance of value below 570000.0
• 40.00% chance of value below 940000.0
• 60.00% chance of value below 1460000.0
• 80.00% chance of value below 2800000.0
• 90.00% chance of value below 4350000.0

Forecaster 2: Probability distribution:

• 10.00% chance of value below 310000.0
• 20.00% chance of value below 515000.0
• 40.00% chance of value below 875000.0
• 60.00% chance of value below 1300000.0
• 80.00% chance of value below 2350000.0
• 90.00% chance of value below 3350000.0

Forecaster 3: Probability distribution:

• 10.00% chance of value below 370000.0
• 20.00% chance of value below 570000.0
• 40.00% chance of value below 930000.0
• 60.00% chance of value below 1500000.0
• 80.00% chance of value below 2850000.0
• 90.00% chance of value below 4200000.0

Research Summary

The research compiles recent NHTSA and news data through May 29, 2026 showing unusually high recall activity in early 2026, driven chiefly by Ford: 34 recall campaigns affecting about 9.8 million vehicles in the first five months of 2026, with the largest single campaign reported at roughly 4.3–4.4 million vehicles (an Integrated Trailer Module/software defect). Q1 2026 total recalls were reported at about 11.6 million vehicles, and other notable 2026 recalls include Hyundai (~421,000 vehicles), Toyota (~43,566), and Tesla (~14,575). The research also cites 2025 totals (~30 million vehicles recalled) and company-level 2025 figures (Ford ~13 million across 153 campaigns).

For historical context and base rates, the research notes that recall-size distributions are heavily skewed (many small recalls, a few very large ones), with the Takata airbag recalls as the historical outlier (~67 million U.S. vehicles, ~100 million global across many years). Recent single-campaign maxima in the 2024–2026 window are in the low millions (e.g., Ford’s ~4.3–4.4M), while typical large campaigns more commonly fall in the hundreds of thousands to low millions; the research presents these as reference-class benchmarks for forecasting the largest potentially_affected value in the June 1–August 11, 2026 window.

Sources used in the research (verbatim links were not provided in the materials I was given; sources cited by name): NHTSA recall database and reports; news reports on Ford 2026 recall activity and the March 2026 4.3–4.4M campaign; 2025 NHTSA/industry recall summaries; reporting on the Takata airbag recalls; news items on Hyundai, Toyota, and Tesla recalls in 2025–2026.

RESEARCH

Report 1 Research

I’ll help you research this NHTSA recall forecasting question. Let me gather relevant news, historical data, and base rates to inform the forecast.

Let me search for prediction markets and more specific information about this forecasting question.

Forecast — Post 370

Modal: 668750-1287500 (30.0%) • partial_shift • 6m2s

Interpretation

Reading: strict

The resolution criteria are highly prescriptive: they name an exact data source URL, exact timestamp for querying, exact date-range filter, exact field name, exact parsing rule, and exact aggregation function (maximum). There is essentially no interpretive latitude — a resolver either follows the steps literally or the question is unresolved.

Time bounds

• observation_window: NHTSA recall campaigns with report_received_date from 2026-06-01T00:00:00.000 through 2026-08-11T23:59:59.999, as recorded in DOT Data Portal Recalls Data at 23:00 UTC on 2026-08-11
• resolution_date: 2026-08-11

Edge cases identified

1. The fixed window starts on 2026-06-01, so campaigns received before June 1 (including May 2026) are excluded even if very large.
2. The question asks for the MAXIMUM single-campaign value, not a sum or count; a single massive recall (e.g. tens of millions) would dominate.
3. If the data source is updated after the 23:00 UTC snapshot (e.g. late filings or corrections posted after that time), those rows would not be captured.
4. If ‘potentially_affected’ contains non-integer or missing values, those rows are excluded — very large recalls with malformed data would not count.
5. The answer range (50,000 to ~5,000,000) implies the resolver expects at least one mid-to-large recall in the window; a value of 0 is technically possible if no campaigns are filed but is outside the stated range.
6. Ties (two campaigns with the same maximum value) do not affect resolution — the MAX value is used regardless.
7. Dataset revisions or retroactive date corrections could shift whether a campaign falls inside or outside the window.

Research (3/3 variants, shared evidence pool)

Total evidence registered (shared pool): ?

Research Brief

Evidence confidence: medium

Scenario 1: Moderate maximum — single-campaign peak in the 500K–2M vehicle range [high evidence]

Conditions favoring

The question window (June 1–August 11, 2026) mirrors the June–August 2025 period, where the largest single campaign was 850,318 vehicles (Ford 25V455000, July 2025). If 2026 summer recall patterns are similar to 2025, the maximum would land in the 500K–1.5M range. The top-10 largest 2025 campaigns all fell between ~444K and ~1.46M (per source 30). No campaign in the comparable June–August 2025 window exceeded ~1.5M (per sources 15, 28). Q2 2025 saw a maximum of ~3.3M Ford vehicles but spread across 49 campaigns, not a single mega-campaign (per source 19).

Conditions against

A single mega-campaign from a large manufacturer (Ford, GM, Tesla, Toyota) could easily push above 2M. Q1 2026 already demonstrated a 4.3M campaign is possible. Q3 2025 saw over 1.75M in a single issue. The arrival model shows 87% probability of at least one >1M campaign, suggesting the top end could exceed 1.5M.

Scenario 2: Large maximum — single-campaign peak in the 2M–5M vehicle range [medium evidence]

Conditions favoring

Q1 2026 saw a 4.38M vehicle campaign (Ford 26C10, filed February 2026), demonstrating that campaigns of this scale occur and may cluster. Q3 2025 saw Ford alone account for 5M+ total vehicles across campaigns (per source 33). Tesla has averaged ~321K per campaign with OTA-resolved issues, but could trigger a very large single campaign (per source 34). The arrival-rate model indicates 87% probability of at least one >1M campaign, with significant right-tail probability of a very large single campaign (per source 39). If a Ford, GM, or Tesla OTA-type campaign triggers in this window, 2M–5M is plausible.

Conditions against

The June–August 2025 comparable window produced a maximum of only ~850K (per source 15). The Q1 2026 mega-campaign was an outlier; 2024 saw no individual campaign approach 4M+ (per sources 11, 34). NHTSA staffing cuts may reduce agency-initiated large recalls (per sources 27, 37). The 2025 full-year maximum was only 1.46M, making 2M+ significantly above the recent trend.

Scenario 3: Very large maximum — ARC Automotive or similarly massive campaign exceeding 5M vehicles [low evidence]

Conditions favoring

NHTSA has been pushing since 2024 toward a recall of ~49–51 million vehicles for ARC Automotive airbag inflators, calling the defect urgent (per source 32). If NHTSA issues this recall order in the June–August 2026 window — either voluntarily by automakers or via forced order — the potentially_affected value could dwarf all other campaigns. The April 2026 precedent of NHTSA’s first-ever forced recall order in decades shows the agency is willing to use this power (per source 38). Additionally, a large OTA software recall (as Tesla has done) could theoretically cover millions in a single campaign.

Conditions against

As of mid-2024 the ARC recall was still being contested by automakers; no evidence of resolution by the research date (per source 32). The April 2026 forced recall covered only a narrow aftermarket set, not ARC Automotive (per source 38). NHTSA staffing cuts reduce agency capacity to force large recalls (per sources 27, 37). The answer range ceiling of ~5M in the question framing suggests the resolver does not expect a >5M outcome. Historical single-campaign maximums in 2025 were capped at ~1.46M; in 2024 Tesla’s full-year total across 16 campaigns was ~5.1M (not a single campaign).

Scenario 4: Below-average maximum — single-campaign peak under 500K vehicles [low evidence]

Conditions favoring

If 2026 summer is a ‘quiet’ recall period following the Q1 2026 mega-campaign spike, regression to the mean could produce a window where no campaign exceeds 500K. Q2 2024 saw only 4.6M total vehicles recalled — significantly below-trend (per source 31). The 10th-largest 2025 campaign was ~444K, suggesting many campaigns fall below this threshold. NHTSA staffing/political headwinds could delay large-scale recalls (per sources 27, 37, 20).

Conditions against

The arrival model projects 87% probability of at least one >1M campaign in the window (per source 39). 2026 recall pace has been elevated vs. prior years (11.6M in Q1 alone, per sources 2, 9). Even Q2 2024’s ‘slow’ quarter saw multiple campaigns above 500K. The comparable June–August 2025 window produced an 850K campaign (per source 15).

Key uncertainties

• Whether any mega-campaign (>2M vehicles) will be filed in the specific June 1–August 11, 2026 window, given that the Q1 2026 Ford mega-campaign was filed before this window opens
• Status of the ARC Automotive airbag inflator recall process — if forced or voluntarily issued in this window, it would be the dominant factor
• Whether NHTSA’s reduced staffing and political environment under the Trump administration significantly dampens voluntary recall filings or agency-initiated campaigns in summer 2026
• Whether large OTA software recalls from Tesla or other OTA-capable manufacturers will occur in this window
• Whether the data.transportation.gov dataset at the 23:00 UTC snapshot will fully capture any recalls filed very close to August 11 (given daily update cadence and 5-business-day filing lag)

Forecast variants (3/3)

Variant 0: inside_view (inside_view_v1)

Inside View: Causal Mechanism Analysis

Current state → decision chain:

The question asks for the MAX single-campaign potentially_affected value in the ~10.6-week window June 1–August 11, 2026.

Key causal anchors:

1. Comparable period base rate (high evidence strength): The directly analogous June–August 2025 window produced a maximum of ~850,318 vehicles (Ford 25V455000, July 7, 2025; per source 15). The top-10 largest 2025 campaigns ranged from 443,899 to 1,456,417 vehicles (per source 30). No single 2025 campaign exceeded ~1.46M. This is the strongest comparator.

2. Arrival-rate model (strong quantitative signal): The arrival forecast (source 39) estimates 87% probability of at least one >1M-vehicle campaign in the window, with median ~4 such campaigns. This strongly supports landing in bin_1 or bin_2 range (668K–1.9M) rather than sub-500K.

3. Q1 2026 mega-campaign (demonstrates scale, but outside window): Ford’s 26C10 (Feb 2026) at 4,380,609 vehicles shows that 4M+ campaigns are achievable (per source 23), but this is before the window opens. The question is whether such a campaign recurs in summer 2026.

4. Medium-strength scenario (2M–5M): Q3 2025 saw ~1.75M in a single backup-camera issue (per sources 29, 33). Ford alone had 5M+ across Q3 2025 campaigns. A repeat is plausible but not the modal outcome for a single campaign.

5. Low-strength scenarios: ARC Automotive (49M+ vehicles) is tagged low evidence — contested as of mid-2024, no evidence of resolution in 2026 window (per source 32). The “quiet” scenario (<500K) is also low evidence given the 87% arrival model for >1M.

Probability weights by bin:

• Bin 0 (below 50K): Near-impossible given any realistic recall activity. ~1%
• Bin 1 (50K–668,750): The “quiet/below-average” scenario. Possible if 2026 summer is unusually slow. ~15%
• Bin 2 (668,750–1,287,500): Centered near the summer 2025 comparable (850K). The high-evidence scenario supports this. ~28%
• Bin 3 (1,287,500–1,906,250): Upper part of 2025 observed range (~1.46M max). Consistent with arrival model’s median result. ~22%
• Bin 4 (1,906,250–2,525,000): Requires a moderately large campaign above 2025’s max. Plausible given Q3 2025 had ~1.75M single-issue (per source 33). ~13%
• Bin 5 (2,525,000–3,143,749): Would require something approaching Ford’s software campaign scale. Below average for 2026 pace but possible. ~7%
• Bin 6 (3,143,749–3,762,499): Getting into the 3M+ territory. Rare but Ford demonstrated 4.3M in Feb 2026. ~4%
• Bin 7 (3,762,499–4,381,249): Approaching the Q1 2026 Ford level. Low probability for a single summer window campaign. ~4%
• Bin 8 (4,381,249–4,999,999): Near or matching the Ford 26C10 scale. Very rare. ~3%
• Bin 9 (above 4,999,999): Would require ARC Automotive or similar extraordinary event. Low evidence, but non-trivial tail. ~3%

The high-evidence scenario anchors bins 2–3 (the 668K–1.9M range), capturing ~50% probability. The medium-evidence large-campaign scenario spreads across bins 4–8. The low-evidence ARC scenario covers bin 9. The low-evidence quiet scenario covers bins 0–1.

Variant 1: outside_view (outside_view_v1)

Reference class: the largest single NHTSA recall campaign observed in a comparable ~10–11 week fixed window (e.g., June–Aug quarters) as recorded in the DOT Recalls dataset. Base rate: an arrival-model calibrated to historical >1M campaigns gives an 87% probability that at least one campaign >1M vehicles appears in a ~10.6-week window (source 39). That implies the maximum in the window exceeds ~1M with ~0.87 base probability. Adjustment: slight downward tilt (–~0.02–0.05) from the pure arrival-model tail to reflect (a) that the Q1 2026 4.38M Ford campaign was an outlier already filed before this window (source 23), (b) the comparable June–August 2025 maximum was ~850K (source 15), and (c) dataset snapshot timing and a 5-business-day filing lag could omit very-late filings at the 23:00 UTC snapshot (sources 40–43). Final distribution (weights follow evidence-strength tags in the brief): I place most mass on a moderate maximum in the ~669k–2.525M range (bins 2–4) because the brief’s “Moderate maximum — 500K–2M” scenario is tagged high evidence (sources 15, 28, 30, 16). Concretely, bins 2–4 together receive 0.69 probability (bin_2 0.09; bin_3 0.35; bin_4 0.25). I allocate 0.23 probability across 2.525M–4.381249M (bins 5–7) to reflect the medium-strength scenario that a mega-campaign (2–5M) could occur (supported by the Feb 2026 Ford 4.38M example — source 23 — and the arrival-model right tail — source 39). A small tail (0.02) is placed on 4.38M–5M (bin_8) and 0.02 on >5M (bin_9) to allow for low-evidence but possible very-large outcomes (ARC airbag or similarly massive forced recall discussed in sources 32 and 38; scenario tagged low evidence). The lower-than-1M mass (bins 0–2 combined = 0.13) equals the arrival-model p(no >1M) ≈ 0.13 and is split with small weight on <50k and 50k–668,750 bins (0.01 and 0.03) and the remainder in 0.09 (bin_2) to reflect historical comparable-window outcomes (June–Aug 2025 peak ≈850k; sources 15, 30). Calibration notes and constraints: - I followed the brief’s scenario evidence-strength tags: high weight to the 500k–2M scenario, medium weight to 2–5M, and low weight to both below-500k and >5M outcomes, per hard rule 5. - I used the arrival-model (source 39) as the principal outside-view quantitative anchor (treated as strong evidence per the research brief guidance). - I accounted for dataset rules (23:00 UTC snapshot, include only rows with report_received_date in 2026-06-01..2026-08-11 and parsable non-negative integer potentially_affected), and for 5-business-day filing lag risk (sources 40–43), which slightly reduces tail mass for very-late filings. - Evidence confidence: medium (per brief). This is why the distribution keeps a noticeable right tail (to reflect observed >1M campaigns historically, e.g., Feb 2026 Ford 4.38M — source 23) but does not concentrate mass beyond ~2.5M without stronger direct evidence. Key citations used in reasoning: arrival-model p(≥1M in window)=0.87 (source 39); June–Aug 2025 largest campaign ≈850,318 (source 15); Feb 2026 Ford 26C10 = 4,380,609 (source 23); dataset update frequency and filing-lag considerations (sources 40–43). Monotone-floor note: resolved value is the running maximum in the fixed window and can only increase as recalls arrive through the window; forecasts therefore anchor on historical arrival rates rather than current-in-window realized maxima.

Variant 2: contrarian_v2 (contrarian_v2)

Adversarial reading

The brief’s bulk of evidence (tagged high-strength) converges on a moderate single-campaign maximum in the 500k–2M range for the June–August 2026 window, mirroring the 850k Ford campaign that was the largest in the directly comparable June–August 2025 period and the top-10 2025 campaigns all falling between 444k–1.46M.

Counter-evidence found (or absence)

The medium-strength scenario for a 2M–5M peak is supported by the documented 4.38M Ford campaign in Q1 2026 and the arrival-rate model’s right tail, but this is explicitly tagged medium rather than high; the low-strength >5M (ARC) scenario rests on 2024 status with no 2026 resolution shown. No high-strength counter-evidence contradicts the moderate-peak consensus.

Distribution implications

High-evidence moderate scenario therefore receives the large majority of mass in bin_1–covering 500k–1.9M (bin_3), with only modest spillover into bin_4–bin_5 from the medium-strength 2M–5M scenario and negligible mass on the low-evidence extremes (bin_0 and bin_9).

Consolidate

Strategy: logit_mean • Drafts: 3/3

Reviewers (3/3)

anchoring → shift_lower (low magnitude, medium confidence) anchoring_v1

The draft forecast’s central mass is reasonably placed, but I detect a mild optimistic anchor in the upper tail. The draft assigns ~17% combined probability to bins 5–9 (above 2.525M vehicles), which is driven by treating the Ford 26C10 Q1 2026 campaign (4.38M vehicles) as evidence that a similar event will recur in the summer 2026 window. However, the brief explicitly notes this campaign was filed before the window opens and represents “the largest single NHTSA recall campaign since a 4.8 million vehicle Chrysler campaign in Q2 2018” — an ~8-year rare event. The draft also leans on the arrival-rate model’s claim of “87% probability of at least one >1M campaign” (source 39) while not adequately discounting that this model tracks campaigns >1M, not >2M; the comparable June–August 2025 window’s actual maximum was only 850K, well below 1M. The modal scenario by direct historical analogy (summer 2025 comparable: 850K peak) places the p50 in 668K–1.29M (bin_2), yet the draft’s p50 sits squarely in 1.29M–1.9M (bin_3), suggesting the center is pulled upward by the outlier Q1 2026 Ford campaign rather than the median of historical summer-window comparables.

Flagged concerns

• Optimistic anchor from pre-window outlier: The Ford 26C10 Q1 2026 campaign (4.38M vehicles, filed Feb 2026) is cited as evidence that 2M–5M campaigns are ‘plausible’ in the summer window. But this campaign was outside the resolution window and is the largest single NHTSA campaign since 2018 — using it as a representative probability weight for bins 6–8 inflates the upper tail inappropriately.
• Arrival-rate model misapplied to upper bins: Source 39’s ‘87% probability of at least one >1M campaign’ is being used to justify both bins 2–3 AND bins 4–9. The model tracks arrivals >1M, not >2M or >4M. The right-tail of that model (p95=16 campaigns >1M) says nothing specific about the probability that any single campaign exceeds 2.5M, which is the threshold for bins 5+. The draft appears to propagate this model’s uncertainty into the high bins without a separate calibration for that threshold.
• p50 shifted above direct historical comparator: The directly comparable June–August 2025 window produced a maximum of 850,318 vehicles. The draft’s p50 (cumulative 50th percentile) falls in 1.29M–1.9M (bin_3), above the 2025 analog. While 2026 may differ, the center should be closer to the 2025 analog absent specific evidence of a known pending mega-campaign in the window.

ceiling → shift_lower (low magnitude, medium confidence) ceiling_v1

The key structural constraint here is that the question asks for the MAXIMUM single-campaign value in a fixed window — this is a running maximum over ~10.6 weeks of incoming recall filings. The brief establishes that in the directly comparable June–August 2025 window, the largest single campaign was 850,318 vehicles (source 15), and the full-year 2025 maximum was 1,456,417 (source 30). The arrival-rate model (source 39) projects 87% probability of at least one >1M campaign, meaning roughly 13% probability the window maximum stays below 1M. The draft places only ~15.6% of mass in bins 0–1 (below 668,750), which is roughly consistent with this constraint. However, the draft somewhat over-weights 1.29M–1.91M (bin_3) at 27.5% given that the 2025 comparable window maximum was 850K and the full-year 2025 maximum was only 1.46M — the constraint implied by the 2025 reference class suggests most mass should concentrate in 668K–1.29M (bin_2), not bin_3. The structural ceiling for “normal” summer activity is approximately 1.5M based on historical data, yet the draft places roughly 35% of mass above 1.9M (bins 4–9), which seems to over-weight the medium/low-evidence scenarios relative to the structural floor set by the 2025 comparable evidence. The draft correctly identifies the >5M (bin_9) tail as very small (~1%), consistent with the ARC Automotive recall being tagged low-evidence.

Flagged concerns

• Structural ceiling from 2025 comparable window: The June–August 2025 comparable window produced a maximum of only 850,318 vehicles (source 15). The draft places only ~22% of mass in 668,750–1,287,500 (bin_2) where this comparable anchor falls, while placing 27.5% in 1,287,500–1,906,250 (bin_3). The structural ceiling from the 2025 comparator implies bin_2 should carry more mass than bin_3, not less. The draft inverts this relationship.
• Right-tail over-weighting relative to structural evidence: The draft places ~35% of mass in bins 4–9 (above 1.9M vehicles). The brief tags the 2M–5M scenario as medium-strength and >5M as low-strength. The 2025 full-year maximum was 1.46M — a structural near-ceiling for ‘normal’ years — making >2M an above-normal outcome. 35% probability for above-normal outcomes in a summer window is generous given the evidence strength tags.
• Insufficient mass near the structural modal outcome: The structural modal outcome implied by the high-evidence scenario (500K–2M, with 2025 comparable at ~850K) spans bins 1–3. The combined mass in bins 1–3 is approximately 63.5%, which is reasonable, but the internal distribution within this range under-weights bin_2 relative to the 850K anchor from the 2025 comparable period.

math → redistribute (medium magnitude, medium confidence) math_v1

The draft distribution exhibits a critical inconsistency between the rationale’s stated scenario weights and the published bins. The rationale explicitly assigns high evidence to the “500K–2M range” scenario, which corresponds to bins_1–3 (50K–1.906M), yet the draft allocates only 0.3896 cumulative probability across these three bins. Meanwhile, bins_2–4 (668.75K–2.525M) receive 0.6628 cumulative probability—a materially different range. The arrival-rate anchor from source 39 (“87% probability of at least one >1M campaign”) should concentrate mass above 1M, but bins_0–1 together receive 0.155 probability, overshooting the ~13% implied by 1 − 0.87. The distribution’s center of mass is shifted upward (mode in bin_3 at 0.275) relative to the high-evidence comparable June–August 2025 maximum of ~850K, which falls in 668.75K–1.287M, assigned only 0.223 (bin_2).

Flagged concerns

• Scenario weight mismatch: Rationale designates ‘Moderate maximum 500K–2M’ as high-evidence scenario (sources 15, 28, 30) and tags it ‘relative_evidence_strength: high’. This spans bins_1–3 (50K–1.906M) = 0.3896 cumulative. Yet the rationale’s own probability weights section assigns bins_2–3 combined 0.28 + 0.22 = 0.50 (excluding the full bin_1), and the draft distribution places bins_2–4 combined at 0.6628—shifting the high-evidence scenario center upward by ~0.3M vehicles relative to the 850K comparable and the stated bin allocations.
• Arrival-model calibration error: Source 39 provides arrival-rate model: p(≥1M campaign in window) = 0.87, so p(<1M) ≈ 0.13. Bins_0–1 (Below 1M) sum to 0.01946 + 0.13613 = 0.155, overshooting the 0.13 implied by the model by +0.025 probability. This suggests either the arrival-model anchor was not fully internalized or the rationale’s earlier statement ‘strong quantitative signal’ was not operationalized consistently into the final map.
• Comparable-window base-rate underweight: The June–August 2025 comparable period (source 15) produced a maximum of ~850,318 vehicles, which falls in 668.75K–1.287M (bin_2). The rationale cites this as the ‘strongest comparator’ yet assigns bin_2 only 0.2235 probability—less mass than bins_3 (0.275) or 0.164 (bin_4). If the high-evidence comparable is truly the strongest anchor, the modal bin should be bin_2, not bin_3.
• Medium vs. high scenario probability split unclear: The rationale states the high-evidence scenario receives ~50% probability and the medium-evidence 2M–5M scenario receives moderate spillover, but the draft shows bins_4–7 (the 2M–5M band) combined at 0.2044, which is only ~32% of the distribution. This is consistent with calling it ‘medium-strength,’ but the published 0.2751 in 1.287–1.906M (bin_3) is not clearly attributed to either scenario in the rationale, creating ambiguity in how the scenarios map to the final distribution.

Reconcile: partial_shift

Reviews consumed: 3

All three reviewers converge on the same structural concern: the draft’s modal bin should be 668,750–1,287,500 (bin_2) rather than bin_3, because the directly comparable June–August 2025 window produced a maximum of ~850K (which falls in bin_2). The anchoring reviewer flagged that the Q1 2026 Ford 26C10 outlier (pre-window) inflated the upper tail; the ceiling reviewer flagged that bin_2 should carry more mass than bin_3 given the structural ceiling implied by the 2025 comparable; and the math reviewer flagged that bins_0–1 overshoot the arrival-model implied ~13% and that bin_2 should be the mode. I accepted the bin_2 > bin_3 ordering and shifted mass from bin_3 down to bin_2, while keeping the upper tail (bins 5–9) only slightly trimmed since the Q1 2026 4.38M precedent and arrival-model right tail remain real signals. I kept bin_9 at 0.04 to preserve some ARC/extreme tail and accommodate the math reviewer’s concern that the right tail was previously underweighted relative to arrival-model uncertainty.

Rejected critiques

• [anchoring] Suggested even larger downward shift of upper-tail mass given Q1 2026 Ford was a pre-window outlier.
  • Why rejected: Partially accepted by trimming bins 5–8 modestly, but I did not fully collapse the upper tail because the 2026 baseline recall pace is elevated vs. 2025 (Q1 2026 saw 11.6M total vehicles recalled) and the arrival-model right tail genuinely supports residual probability of a 2M+ campaign.

Final Distribution

Evidence Sources (44)

1. Over 12 Million Vehicles Recalled in Q1 2026 - BizzyCar — Ford’s 26C10 electrical system recall in Q1 2026 impacted over 4.3 million vehicles, accounting for more than 35% of all vehicles recalled in Q1 2026.
2. ABC4 Utah Facebook post on Q1 2026 recalls — In the first three months of 2026, more than 11.6 million vehicles were recalled in the U.S., including a large GM recall covering Chevrolet Silverado and GMC Sierra 1500, 2500 and 3500 pickups from 2020-2022 model years.
3. Vehicle Recalls: What 2024’s totals mean for drivers in 2025 - Motus — In 2024, just over 27.7 million vehicles were recalled in total by NHTSA, down from 33.6 million in 2023 and 29.0 million in 2022.
4. Automotive Recall Alert: BizzyCar Q1 2025 Recall Report — Q1 2025 saw 3.46 million vehicles recalled, described as the lowest quarterly total in over 10 years.
5. NHTSA Vehicle Safety Recalls Week — In 2025, NHTSA documented more than 29 million vehicles recalled total (across 997 recalls for vehicles, car seats, tires, RVs and other equipment).
6. Automotive Recalls in 2025: A Defining Conquest Strategy — In Q2 2025, more than 7.3 million vehicles were recalled across 49 OEM campaigns, with Ford, GM, Stellantis, Nissan, Honda, and Toyota leading.
7. Recalls Were Up in 2025, but So Were Completed Repairs: State of Recalls Report — In 2025, 447 NHTSA-mandated and 223 voluntary recall campaigns affected over 28 million vehicles (670 total campaigns), per State of Recalls report.
8. NHTSA Report on Vehicle Safety Recall Completion Rates (Jan 2025) — A 2025 NHTSA report on recall completion rates noted the largest single recall had over 5 million vehicles affected, and one recall involving more than 1 million vehicles occurred between 2010 and 2022.
9. 11.6 Million Vehicles Recalled in 2026: What You Must Know — In Q1 2026, more than 11.6 million vehicles landed on recall lists in the United States across three months; the 2026 recall crisis has included recalls of 1.7 million Ford vehicles with failing backup cameras and nearly 5 million affected by braking issues.
10. NHTSA 2024 Annual Report - Safety Recalls — NHTSA’s 2024 Annual Recall Report (published April 2025) shows annual vehicle recall counts from 2004-2024, with 2021 seeing 1,093 vehicle safety recalls affecting 22.8 million vehicles — the highest number in a single year historically reported.
11. Most recalled automaker of 2024? It wasn’t Ford - CBS 17 — In 2024, Ford had 67 recalls totaling 4,777,123 vehicles. The largest single recall involved door trim pieces detaching from older Explorers.
12. U.S. Vehicle Recalls Hit Multi-Year High in Q1 2026 - PR Newswire — In Q1 2026, U.S. vehicle recalls hit a multi-year high; a single Ford electrical system recall impacted more than 4.3 million vehicles, accounting for nearly 40% of all Q1 2026 recall volume.
13. Ford Recalls 4 Million Vehicles Over Software Glitch - Forbes — Ford recalled more than 4 million vehicles (including F-series trucks and SUVs) over a software glitch, filed in February 2026 — this is the largest single NHTSA campaign identified in early 2026.
14. Ford’s record 90 NHTSA recalls in 2025 - Recall Masters via LinkedIn — In Q3 2025, NHTSA saw 8.5 million vehicles recalled, the highest quarterly total in nearly two years, according to Recall Masters analysis.
15. Ford recalls 850k vehicles: Mexico supplier fuel pump mishaps — The largest single recall in the June–August 2025 period was Ford NHTSA Campaign 25V455000 (fuel system, gasoline) affecting 850,318 vehicles (announced July 2025), covering 2021-2023 Bronco, Explorer, F-150, F-250/350/450/550, Lincoln Aviator, Navigator, Mustang, and Expedition.
16. Insider Trading Around Auto Recalls: Does Attentiveness Matter? (SSRN) — Academic research found the average number of cars potentially affected per NHTSA recall campaign is approximately 226,209 vehicles (based on historical data through study publication date).
17. Recall Rundown: August 2025 - RecallMasters — In August 2025, Ford recalled 850,318 certain 2021-2023 Bronco, Explorer, Lincoln Aviator, and F-series vehicles — one of the larger single campaigns in summer 2025.
18. 2026 Auto Recalls: Ford Cameras, Toyota Highlander Seats & GM - SamAndAsh Law — As of May 2026, Ford leads with over 7.3 million vehicles affected across 17 campaigns in 2026 to date; Toyota, GM, Hyundai, and others also had significant 2026 recall campaigns.
19. Automotive Recall Alert: Over 7.3 Million Vehicles Recalled in Q2 2025 - BizzyCar — In Q2 2025 (April–June), over 7.3 million vehicles were recalled — the highest quarterly total since Q1 2024’s 9.9 million. Ford alone accounted for 3,322,450 vehicles across 49 campaigns (45.3% of Q2 total). The leading recalled component was back-over prevention affecting 2.59 million vehicles.
20. NHTSA Suspends Recall Updates During Federal Shutdown — NHTSA suspended recall updates during a federal shutdown period in 2025/2026, pausing regulation of vehicle recalls and not posting updates on its website during the shutdown.
21. Recall Rundown: August 2025 - Recall Masters — In August 2025 alone, there were 90 NHTSA recalls affecting almost 6 million vehicles, suggesting an average of roughly 67,000 vehicles per recall campaign in that month.
22. Carfax Recall Tracker for May 2026 — NHTSA Campaign 26V237000 (received 4/14/26) has a potential number of vehicles affected of 1,392,935 — one of the larger single-campaign recalls in early-to-mid 2026.
23. Ford Recalls 4.4 Million Trucks and SUVs - CarEdge — The Ford 26C10 campaign (filed February 2026) covered 4,380,609 vehicles — 2021-2026 F-150, 2022-2026 Super Duty, Maverick, Expedition, Navigator, and 2024-2026 Ranger — for a software bug in the Integrated Trailer Module disabling trailer brakes/lights. This is the largest single recall campaign since a 4.8 million vehicle Chrysler campaign in Q2 2018.
24. BizzyCar Q1 2026 Recall Report — Q1 2026 recall activity was highly concentrated: a single Ford electrical system campaign accounted for ~40% of all vehicles recalled that quarter. The pattern shows recalls becoming “less predictable and more event-driven,” with individual campaigns capable of reshaping the overall landscape.
25. Ford Is The Most Recalled Brand Of 2025 - CarBuzz — The single largest NHTSA recall in 2025, by volume, was a September 4, 2025 notice for 1,456,417 Fords and Lincolns with blank or distorted rearview cameras (per CarBuzz analysis).
26. BizzyCar Q1 2026 Recall Report — Ford’s 26C10 campaign alone affected 4.3 million vehicles in Q1 2026, making it the largest single NHTSA recall campaign found in 2026 data through Q1 (received before June 1, 2026).
27. US auto safety agency shedding more than 25% of employees - Reuters — NHTSA lost 25% or more of its staff due to DOGE cuts and buyouts as of mid-2025; as of July 2025, the agency was shedding more than 25% of employees, with staffing subject to further cuts if Congress reduces the budget.
28. 10 Biggest Recalls of 2025 - CarBuzz — The 10th-biggest recall of 2025 affected 443,899 units (Nissan/Infiniti engine failure recall 25V437000, issued June 27, 2025), and the 5th-biggest affected 850,318 units (Ford fuel pump, Recall 25V455000, issued July 7, 2025). Thus the top-10 recalls in 2025 ranged from ~444K to ~1.46M vehicles.
29. Automotive Recall Alert: 8.5 Million Vehicles Recalled in Q3 2025 - BizzyCar — In Q3 2025 (July-September), automakers issued 96 recall campaigns covering 8,494,262 vehicles total — a 16% increase from Q2’s 7.32 million. This is the most relevant comparable period to the question’s June-August 2026 window.
30. 10 Biggest Recalls of 2025 - CarBuzz — In 2025, CarBuzz confirmed the top-10 recalls by volume (based on NHTSA data through Dec 12, 2025) ranged from 443,899 to 1,456,417 units per campaign, with #2 at 1,448,655, #3 at 1,076,138, #4 at 1,024,407 (Toyota/Lexus), #5 at 850,318 (Ford July 7), #6 at 591,377, #7 at 597,571 (GM), #8 at 568,580 (Hyundai), #9 at 499,129 (Ford Aug 22), #10 at 443,899 (Nissan June 27). No single campaign exceeded ~1.46M.
31. Automotive Recall Alert: Over 14 Million Vehicles Affected Year-to-Date 2024 - BizzyCar — In Q2 2024 (April–June), only 4,621,994 vehicles were recalled across 115 campaigns — significantly less than Q1 2024’s 9.9 million. This shows substantial quarter-to-quarter variability in recall volumes; a “post-spike” quarter can be much lower.
32. US stands by decision that 50 million air bag inflators are dangerous - KCTV5 — As of July 2024, NHTSA was pushing toward a recall of 51 million ARC Automotive airbag inflators in ~49 million vehicles. If this recall is ultimately forced, it could appear in 2026 and dominate a single quarter’s largest-campaign metric. However, automakers opposed the recall and the process was ongoing as of mid-2024.
33. Vehicle recalls surge in Q3 for highest quarterly total in almost two years - Dealership Guy News — In Q3 2025, Ford led all manufacturers with 5,041,241 vehicles affected across campaigns tied to ~5 million units; the single largest Q3 2025 issue was backup camera failures affecting over 1.75 million vehicles.
34. Tesla leads the 2024 recall list, followed by these automakers - Autoblog — Tesla recalled the most vehicles of any single manufacturer in 2024 at 5,135,991 total (across only 16 campaigns, so average ~321K per campaign). The vast majority (99.2%) were via OTA software updates. In 2024, Tesla’s average recall size per campaign was very large due to OTA-resolved software issues.
35. Recall Rundown: April 2026 - Recall Masters — In April 2026, Ford recalled 889,950 vehicles due to problems with a rearview camera that may fail to display images correctly, per Recall Masters April 2026 Recall Rundown.
36. Over 12 Million Vehicles Recalled in Q1 2026 - BizzyCar — In Q1 2026, five manufacturers (Ford, Toyota ~1.0M, Hyundai ~806K, Chrysler ~726K, Nissan ~669K) accounted for 93.7% of all vehicles recalled, with Ford’s single campaign (26C10) alone covering 4.3M vehicles.
37. As Recalls Slow & Rulemaking Stalls, What’s Next for the Nation’s Vehicle Safety Watchdog? - Reinhart Law — Under the Trump 2.0 administration (Jan 2025 onward), NHTSA’s ODI investigations slowed: only 4 new investigations were initiated Jan–March 2025 (vs. 19 in the same quarter in 2024), and recall counts also fell from 271 in FY25Q1 to 222 in FY25Q2. However, voluntary manufacturer-initiated recalls (which dominate large-scale campaigns) are not directly controlled by NHTSA staffing levels.
38. NHTSA Issues the First Defect Recall Order in Decades - Sidley Austin — On April 29, 2026, NHTSA issued a first-ever recall order in decades — for Chinese-made aftermarket airbag inflators (DTN brand) that caused deaths and injuries. This recall order did NOT cover tens of millions of vehicles but targeted aftermarket equipment. This was NOT the ARC Automotive recall.
39. Arrival forecast model for large NHTSA recall campaigns — Arrival forecast (rate: ~0.5 recall campaigns per week with >1M vehicles affected, over ~10.6 weeks, overdispersion 0.8) yields: mean ~5.3 such campaigns, p5=0, p25=1, median=4, p75=8, p95=16. This suggests 87% probability of at least 1 mega-campaign (>1M vehicles) in the June 1–August 11 window, supporting the expectation of a large single-campaign maximum.
40. NHTSA’s Office of Defects Investigation (ODI) - Recalls - NHTSA API | DOT Data Portal — The NHTSA ODI Recalls – NHTSA API dataset on the DOT Data Portal (data.transportation.gov, dataset ID aqh3-3rri) lists an Update Frequency of “R/P1D” (daily), meaning the dataset is updated on a daily cadence, not in real-time.
41. NHTSA’s Office of Defects Investigation (ODI) - Recalls - Recalls Lookup by VIN | DOT Data Portal — The NHTSA ODI Recalls – Recalls Lookup by VIN dataset on the DOT Data Portal (dataset ID pin3-udpr) also shows Update Frequency “R/P1D” (daily), consistent with a daily rather than real-time posting schedule.
42. NHTSA’s Office of Defects Investigation (ODI) - Recalls - Recalls - Tires | DOT Data Portal — The NHTSA ODI Recalls – Recalls Tires dataset on the DOT Data Portal (dataset ID b6a2-cuwf) lists Update Frequency “R/P1D” (daily), further confirming the recalls data family is refreshed daily.
43. Recalls Data - Catalog (data.gov / DOT Socrata) — Manufacturers are legally required to notify NHTSA within 5 business days of determining a safety defect or noncompliance under 49 CFR Part 573; this filing lag means recalls decided close to the snapshot date of 2026-08-11 23:00 UTC may not yet appear in the daily-updated dataset at the time of the snapshot.
44. Recalls Data - Catalog - Data.gov — The Recalls Data dataset on the DOT Data Portal (identifier 6axg-epim, harvested from DOT Socrata) shows “Data Last Modified: April 6, 2026” and “Metadata Updated Date: April 6, 2026”, confirming the dataset is actively maintained and has been updated as recently as April 2026.

wooley_rhino • run e67421f5 • $3.79 • 6m2s

Extracted Percentile Values: {0.1: 60000, 1.0: 130000, 5.0: 250000, 10.0: 350000, 15.0: 430000, 20.0: 500000, 25.0: 560000, 30.0: 620000, 35.0: 680000, 40.0: 740000, 45.0: 800000, 46.0: 815000, 47.0: 830000, 48.0: 850000, 49.0: 870000, 50.0: 890000, 51.0: 915000, 52.0: 940000, 53.0: 970000, 54.0: 1000000, 55.0: 1030000, 60.0: 1150000, 65.0: 1290000, 70.0: 1440000, 75.0: 1600000, 80.0: 1800000, 85.0: 2050000, 90.0: 2400000, 95.0: 3200000, 99.0: 5200000, 99.9: 9000000}

Model’s Answer:

Analysis of NHTSA Maximum Recall Campaign Size (June 1 - August 11, 2026)

Question Understanding

I need to forecast the largest single potentially_affected value from any NHTSA recall campaign with a report_received_date in the June 1 - August 11, 2026 window (approximately 10.5 weeks).

Historical Benchmarks and Recent Data

2025 Full-Year Data:

• Over 24.4 million vehicles recalled across 360+ campaigns
• Ford’s largest single campaign: ~1.46 million vehicles
• Several campaigns in the 850,000+ range
• Many campaigns in the 100k-500k range

Early 2026 Activity (Elevated):

• Q1 2026 saw >11.6 million vehicles recalled (among highest quarterly totals)
• Reports indicate 2026 recall activity is “three times higher” than 2025 pace
• May 2026 saw Hyundai recall of 421,078 vehicles (software braking issue)
• Toyota recall of 43,566 vehicles (engine issues)
• Ford already issued 840,000+ vehicle recall (fuel pump) and 270,000+ (rollaway risk) in early 2026

Historical Large Single Campaigns:

• Ford ~1.5 million (rear-view camera, Sept 2025)
• Ford 840,000+ (Bronco/Edge, early 2026)
• Hyundai 421,000+ (braking software, May 2026)

Key Factors Affecting Forecast

Why outcome could be LOW (~200,000-400,000):

• No mega-recall happens to fall within this specific 10.5-week window
• Large recalls are filed just before or after the window
• Probability: ~10-15%

Why outcome could be HIGH (~1.5M-3M+):

• Multi-model, multi-year defect discovered affecting major manufacturer
• Software/ADAS issue affecting entire fleet (Tesla, Ford, etc.)
• Continued elevated recall activity yields a mega-campaign
• Probability: ~20-30%

Critical Assumptions and Their Failure Probabilities

Assumption 1: Current elevated recall pace continues through summer 2026

• Failure probability: ~15%
• Impact if wrong: Max could be lower by 200-300k

Assumption 2: Distribution of recall sizes follows historical patterns

• Failure probability: ~10%
• Impact if wrong: Could see unprecedented mega-recall (5M+) or unusual quiet period

Assumption 3: No regulatory changes dramatically affecting reporting

• Failure probability: ~5%
• Impact if wrong: Could affect what gets classified as recalls

Distribution Construction

Given:

• Median expectations around 850,000-950,000 (reflecting typical large campaign sizes)
• Strong upward tail possibility due to elevated 2026 activity
• Limited downside given current recall environment
• Historical ceiling around 1.5-2M for single campaigns, but could exceed in exceptional cases

The 10.5-week window means:

• High probability of at least one 500k+ campaign
• Moderate probability (40-50%) of a 1M+ campaign
• Lower probability (15-25%) of 1.5M+ campaign
• Small probability (<5%) of 3M+ campaign

Final Rationale:

Given the elevated 2026 recall environment (3x 2025 pace), the recent large campaigns from major manufacturers, and historical data showing single campaigns routinely reaching 800k-1.5M, I place my median around 900,000 vehicles. The distribution is right-skewed because mega-recalls can suddenly emerge from software defects or multi-model issues, while there’s a natural floor given that major OEMs consistently file substantial campaigns. The lower bound reflects an unusually quiet period (unlikely given current trends), while the upper bound accounts for rare but possible mega-recalls affecting millions of vehicles.

Percentile .1: 60000 Percentile 1: 130000 Percentile 5: 250000 Percentile 10: 350000 Percentile 15: 430000 Percentile 20: 500000 Percentile 25: 560000 Percentile 30: 620000 Percentile 35: 680000 Percentile 40: 740000 Percentile 45: 800000 Percentile 46: 815000 Percentile 47: 830000 Percentile 48: 850000 Percentile 49: 870000 Percentile 50: 890000 Percentile 51: 915000 Percentile 52: 940000 Percentile 53: 970000 Percentile 54: 1000000 Percentile 55: 1030000 Percentile 60: 1150000 Percentile 65: 1290000 Percentile 70: 1440000 Percentile 75: 1600000 Percentile 80: 1800000 Percentile 85: 2050000 Percentile 90: 2400000 Percentile 95: 3200000 Percentile 99: 5200000 Percentile 99.9: 9000000

Fleet scale, product diversity, and multi-decade campaign records define the feasible magnitude of single-campaign counts.

Regulatory filing cadence and concentrated manufacturer and supplier exposure concentrate risk into discrete calendar windows.

Recent history shows larger per-window maxima and more frequent supplier-led large campaigns, which shifts the empirical center upward.

Ongoing regulatory scrutiny and industry consolidation favor large, consolidated filings rather than many small ones.

A supplier-wide, platform-wide, or software-enabled defect is the primary inflection that can push the window maximum past multi‑million thresholds.

Clustering of filing dates, revisions to reported counts, and late submissions mean the realized maximum can move by millions after initial entries.

Key unresolved uncertainties are unfiled or delayed filings, evolving definitions of ‘potentially affected’, and a sparse historical tail for >5M events.

Sensitivity is highest to whether a single large supplier campaign occurs in the window and whether the modern upward trend in campaign sizes persists.

Forecast Summary

Median: 1,720,858 vehicles

Rationale

• Tesla FSD investigation is the primary upside driver. NHTSA escalated a probe covering ~3.2M Tesla vehicles to engineering analysis (the final pre-recall stage) in March 2026. If converted to a recall during the June–August window, this would be the largest outcome. However, political uncertainty and the non-deterministic nature of the engineering-analysis-to-recall pathway suggest ~20–25% probability of filing within the window.

• Historical baseline is robust. The 11-year record of annual maxima in this exact window (June 1–August 11, 2015–2025) ranges from 505K to 3.62M, with a median of ~1.85M and mean of ~1.96M. This provides a strong structural anchor, and the distribution is approximately lognormal with a heavy right tail.

• Ford’s largest 2026 recalls are already filed. Ford’s four biggest 2026 recalls (4.38M, 1.39M, 1.74M, and others) all occurred before June 1, partially depleting the pipeline. Continued Ford activity is expected but likely at smaller scale. The Stellantis 7.4M probe was closed without recall.

• Moderate-to-large single recall is most likely. The most probable scenario (~37% weight) is a single large platform recall from Honda, Toyota, or GM in the 1–2M range, consistent with 2019, 2023, and 2024 patterns. Honda has historically produced the largest Jun–Aug recall in 3 of 11 years, though no specific investigation is currently identified.

• Uncharacterized NHTSA docket creates tail risk. The research explicitly notes gaps in the full list of active engineering analyses. Unknown investigations nearing recall stage could push the outcome higher, particularly in the 2–3.5M