Introduction
Every deployed public-key scheme rests on a conjecture that some problem is hard. We rarely ask how hard anyone has actually tried. This page reports a literature-based estimate of public cryptanalytic effort for each assumption class behind standardized encryption: the elliptic-curve discrete logarithm (ECDLP), structured and plain lattices (MLWE, LWE), and syndrome decoding, with factoring, isogenies and multivariate systems as comparators.All figures count public work only. Classified effort is discussed in §6.
The goal is a defensible order-of-magnitude comparison of the form class X has received about N times the scrutiny of class Y, not a security level. Computational records and model-based work estimates get their own track.
The PKI and protocol upgrade analysis follows these assumptions into deployed systems. It charts how many CA keys would need attention on the date of an RSA break, tracks RSA and ECDSA in recent issuance, and measures what larger keys and signatures add to a handshake.
A research-allocation year (called research-FTE in the underlying data) is an estimate based on the share of a researcher's publications devoted to a topic. It is not a recorded year of work.
The literature
We harvested IACR ePrint, DBLP, OpenAlex and zbMATH, screened by keyword, and classified candidates against a written rubric. A second pass reviewed included and borderline items plus a sample of exclusions. Figure 1 shows the result by year of publication; papers that address more than one assumption are split fractionally. Every counted item is listed, by year and with BibTeX, in the bibliography.
Lattice research exceeds ECDLP in the main cumulative views. The lattice/ECDLP ratio ranges roughly from 1.3 to 2.6 across the study's counting and source assumptions.
These are different measures and sensitivity choices, not a confidence interval. The main paper-count view crosses in 2017; the author-output estimate crosses in 2018.
Who does the work
Paper counts treat a two-page note and a decade-long program alike. The author map groups 7,303 authorships into about 3,112 identities, some unresolved, and estimates each person's share of publications devoted to a class over a moving five-year window. Two people with half their output in a class contribute one allocation year.
These estimates may substantially overstate researcher time. Publishing only on a topic does not mean working on it full time. Teaching, administration, other employment, short projects and unrecorded work can all make publication share exceed time share. A minimum assumed publication rate limits the effect of sparse records, but does not measure hours. The alternative career-adjusted view reduces shares by assumed research-time fractions; it is also a proxy, not a correction validated against researchers' time records. Some authors lack career data and use assumed effort per paper instead.
Cumulative effort
In this model, factoring and FF-DLP together still lead on accumulated allocation years, while lattices passed ECDLP in 2018. These curves describe the model's allocation of published output over time.
The baseline lattice/ECDLP allocation ratio is 1.79. Across 81 combinations of assumptions about publication rates, career windows, fallback effort and journal extensions, it ranges from 1.70 to 1.93. The year when lattices overtake ECDLP and stay ahead ranges from 2017 to 2020. A separate scope test drops every paper tagged generic lattice or SIS, including papers with other tags. The ratio then falls to 0.55. The white paper explains this deliberately narrow comparison. The scenario results give the totals under each set of assumptions.
Individual researchers
A separate page breaks the published output down by person: who contributed most to each assumption class in any period, how concentrated each field is, and how its researchers are connected through co-authorship.
Researcher explorer Top contributors, concentration and co-author networks by field and periodDistance to the frontier
Effort is one axis; demonstrated computation is another. For each target we use a cost model to estimate the work ratio between an attack and a relevant public computational record. We express this ratio in bits:
A gap of 10 bits means about a thousand times the modeled work; 35 bits means about 35 billion times. Both endpoints use the same model. Only a shared, size-independent cost factor cancels; memory, communication and other costs can change with problem size.
These gaps are not security levels or forecasts of attack cost. Different families use different operations, memory requirements and model assumptions, so similar numerical gaps need not mean similar practical distances. RSA-2048's roughly 35-bit gap is a GNFS extrapolation, not a measured runtime ratio.
The P-256 comparison uses generic group operations. A 29 October paper by Akshima and coauthors gives a conditional subexponential bound for prime-field ECDLP using a proposed tensor-contraction theorem. The authors give no improved concrete cost for P-256.
See the normalization method for the models, sources and verification limits.
Attack history and stability
How have attack algorithms changed? The plot below traces selected, source-checked mathematical bounds. Each formula gives a leading-term index: a simplified measure that drops lower-order factors. A fall in this index records an improvement in the selected model. It does not measure security bits lost, and a flat line does not mean that practical attacks stopped improving.
The first view holds a reference size fixed. The second finds the size giving an index of 128 under each selected bound, shown relative to the latest such size. These are illustrative model sizes, not recommended cryptographic parameters. Dates follow the quoted analyses, including later corrections, rather than reconstructing what researchers believed at each moment.
An unchanged leading exponent is a narrow kind of stability. Better reduction methods, dimensions-for-free, memory tradeoffs and implementations can improve finite-size attacks without changing that exponent. These histories cannot predict future breaks or convert publication counts into a security guarantee.
Two post-cutoff papers report improved worst-case bounds for exact SVP on arbitrary lattices. These guarantees address a different question from the heuristic sieving cost plotted here. See Gao, Han and Hu and Hhan, Kim, Kim and Laarhoven.
Breaks and reassessments with different cost models
Finite-field advances and the SIDH break are retained as sourced events. Their concrete estimates and runtimes concern particular instances and units, so they are not joined to the index curves. Polynomial and quasi-polynomial algorithms are distinct; neither label alone supplies a numerical parameter recommendation.
Rules, exclusions and corrections: detailed methodology. Full dated ledger and sources: algorithmic history report. Sources were checked by an agent; independent human verification remains outstanding.
What this does not count
Our literature estimates cover publicly documented work: papers, preprints, theses and published records. Unpublished failures, private research and classified work are missing. We do not have a reliable allocation of that missing effort by assumption class.
If public effort were known, additional classified effort could be represented by an unknown multiplier:
We do not know whether this multiplier is similar across classes. Comparisons of the visible literature remain comparisons of public work; extending them to total effort requires additional assumptions. Missing work does not make our publication-based time estimates lower bounds: those estimates may themselves be too high.
Why \(k_X\) is not estimated. One might hope to read classified headcount off public government budget documents, taking the NSA as the best-documented case. Official documents do not allow it. The National Intelligence Program is published only as a topline; every NSA line beneath it is classified, the one unclassified NSA budget line covers the defensive mission, and the agency is excluded from the federal workforce statistics that would otherwise count its mathematicians. The leaked FY2013 budget justification is the only document that pairs a dollar line with people: “Cryptologic Math & IDA Research” at $119–146M a year, and one cryptanalysis project with 240 civilian staff on $39M. Dividing the research line by a plausible loaded cost gives roughly 300–500 full-time staff, uncertain by a factor of two, for a budget line rather than for cryptanalysts, as of 2012, with no split by assumption class. An official series does exist for the IDA cryptologic centers that work under NSA contract, and it suggests that such dollar conversions overstate research staff by about a factor of two. The methodology page lists the sources and what each can and cannot bound.
Effort in inference dollars
Recent AI systems have produced new results in mathematics and in cryptanalysis at disclosed costs. We do not estimate an exchange rate between human and machine research. Instead we express historical effort \(H_X\) as an equivalent inference budget under an assumed productivity multiplier \(s\):
The multiplier means assumed research productivity per dollar: at 10×, one estimated human research-year corresponds to $25,000 of inference. No multiplier is measured or preferred. These budgets inherit any overestimate in the human-effort proxy: if actual time is half our estimate, the corresponding budgets are also half as large.
Read in reverse, a $100,000 inference budget represents 0.4, 4 or 40 assumed human-year equivalents at 1×, 10× or 100×. These are assigned equivalents, not observed labor displaced by AI. The paper separately explores HAWK normalization using paper-based effort on both sides; it does not apply that rate to the author-output estimates shown here.
On October 29, Sam Altman estimated the cost of OpenAI’s Navier–Stokes work at roughly $10 million without specifying which expenses that figure covers. Anthropic reports using 14 million output tokens to discover the core algorithm in its Exponential Time Hypothesis (ETH) result. OpenAI’s separate ETH result has no discovery budget in the sources we checked. The disclosure notes give the sources and the scope of each result.
Recent mathematical results
For selected mathematical problems and HAWK, we compare disclosed AI budgets with a paper-based estimate of effort in the surrounding literature. This second estimator assigns assumed years to each paper, rather than reconstructing author publication shares. The samples include partial results and related problem variants.
Long-standing does not necessarily mean hard-fought. An Erdős problem can remain open for decades because few people have seriously tried to solve that particular question. A low-cost AI solution may therefore reveal an overlooked opportunity, rather than overcome decades of concentrated human effort. Some problems have attracted substantial work; others have not, and we do not know the effort spent on unsuccessful attempts. Papers on a surrounding topic may also devote little time to the exact question. Our literature estimates cannot resolve that distinction. These selected successes should not be treated as evidence that AI can cheaply replace an equally large body of sustained cryptanalytic scrutiny.
The literature totals are not yet audited pre-discovery human labor. The retained corpus includes explicitly AI-assisted papers, marked in the chart, and its historical cutoffs still need review. Existing literature also supplies tools to the AI. Comparing these dollar amounts does not measure productivity gains or the fraction of historical scrutiny reproduced.
Sources: FrontierMath Erdős, Tables 2–3 and Anthropic's HAWK report. The Erdős paper reports more than $220,000 across its wider campaign; the individual successful-run costs shown here are a different accounting boundary. HAWK's roughly $100,000 is API cost for discovery, not the cost of executing a key-recovery attack.
Method and caveats
For the exact rules, formulas, source files and unresolved defects, read the detailed methodology and audit trail.
Corpus. IACR ePrint and DBLP harvests, OpenAlex phrase searches and citation expansion, and zbMATH. The October 27 correction of the frozen October 9 corpus contains 4,348 included or borderline records; 2,876 are included items in the main effort categories. With 103 duplicate versions at weight zero and 9 journal extensions at one third, 2,773 records have positive work weight. The correction removes eleven additional duplicate records, repairs an earlier false merge, and excludes one HAWK-only paper under the existing rubric; original model labels are preserved. It adds no new literature scan. Recall against twelve selected seed papers' reference lists is approximately 96–100% per class; this does not establish field-wide recall.
Classification. The second model pass reviewed included and borderline items and a sample
of exclusions, with access to the first labels. Agreement of \(\kappa = 0.84\)–\(0.89\) describes consistency
on the reviewed subset. Exclusion reversal rates of 0.5–1.1% concern sampled first-pass exclusions.
Neither check establishes independent validation, classifier recall or a global error rate.
An independent corpus audit reproduces the counts and provides a 319-record sample in 27 strata for human review
(analysis/corpus_audit_20261027.md). That review remains unfilled; no validated precision estimate is available.
Scope. The rubric excludes attacks whose contribution is extracting power, timing or fault leakage, and implementation or protocol flaws unrelated to the mathematical problem. It intentionally includes mathematical attacks given partial information, quantum algorithms, special or weak instances, and faster implementations of mathematical solvers. These broader categories do not all transfer to ordinary deployed parameters. Abstract-based classification can also admit out-of-scope work or misattribute mixed papers; such errors can overstate a class's relevant scrutiny. See the classification rubric.
Time. Output share may substantially exceed time share. Missing career publications, uncertain identities and effort-per-paper assumptions add further error. Career-stage adjustment lowers the principal-class totals by roughly a quarter, but relies on assumed research fractions and often estimated career stages. No defensible numerical error bar is available. Dollar scenarios scale directly with these uncertain totals.
Models. Pipeline, analysis and text by Claude Fable 5.1 and Claude Opus 5.5, with contributions from GPT-6 Astra; smaller Claude models ran the first classification pass. Methodological decisions were made and recorded by the overseeing human.
Data and citation
Read the detailed methodology for the audit trail and the working white paper (PDF) for the synthesis, nominal person-hours, robustness results, computational records and AI research budgets.
Every figure on this page is drawn from the generated data. The papers behind the
counts are listed in the bibliography (BibTeX). The
study repository contains the sources,
classification rubric and pipeline/build_site_data.py used to regenerate it.
Corrections and suggestions. We would like to hear about anything that looks wrong: a paper that is missing, miscounted or in the wrong class, an author credited to the wrong person, a number that does not match its source, or an assumption in the method that you think is mistaken. Proposals for better methods are just as welcome. Please open an issue on the study repository; the templates ask for the details we need to check a report. Accepted changes are recorded in the repository history, and the site is regenerated from it.
@misc{bitsofwork2026,
title = {How Hard Have We Looked? Measuring Public Cryptanalytic Effort},
author = {Green, Matthew},
note = {AI-conducted analysis, overseen by Matthew Green. Working draft},
howpublished = {\url{https://bitsofwork.org}},
year = {2026},
}