MétaCan
Menu
Back to cohort

The AI Hardness of CAPTCHAs does not imply Robust Network Security

2007· book-chapter· en· W24291140 on OpenAlexaff
Urs Hengartner

Bibliographic record

Venuenot available
Typebook-chapter
Languageen
FieldComputer Science
TopicUser Authentication and Security Systems
Canadian institutionsUniversity of Waterloo
Fundersnot available
KeywordsCAPTCHAComputer scienceComputer securityThe InternetProtocol (science)ArchitectureWorld Wide Web

Abstract

fetched live from OpenAlex

A CAPTCHA is a special kind of AI hard test to prevent bots from logging into computer systems. We define an AI hard test to be a problem which is intractable for a computer to solve as a matter of general consensus of the AI community. On the Internet, CAPTCHAs are typically used to prevent bots from signing up for illegitimate e-mail accounts or to prevent ticket scalping on e-commerce web sites. We have found that a popular and distributed architecture for implementing CAPTCHAs used on the Internet has a flawed protocol. Consequently, the security that the CAPTCHA ought to provide does not work and is ineffective at keeping bots out. This paper discusses the flaw in the distributed architecture’s protocol. We propose an improved protocol while keeping the current architecture intact. We implemented a bot, which is 100% effective at breaking CAPTCHAs that use this flawed protocol. Furthermore, our implementation of the improved protocol proves that it is not vulnerable to attack. We use two popular web sites, tickets.com and youtube.com, to demonstrate our point.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.

How this classification was reachedexpand

Full frame machine prediction

Teacher imitation

Not calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.

metaresearch head score (Codex)0.001
metaresearch head score (Gemma)0.010
Version: metacan-v3-hybrid-931329e0061cValidation status: machine_predicted_unvalidated
Candidate categoriesnone
Consensus categoriesnone
DomainCandidate signal: none · Consensus signal: none
Study designCandidate signal: Theoretical or conceptual · Consensus signal: Theoretical or conceptual
GenreCandidate signal: Methods · Consensus signal: Methods
Teacher disagreement score0.010
Threshold uncertainty score0.032

Distilled classifier scores by category (both heads)

CategoryCodexGemma
Metaresearch0.0010.010
Meta-epidemiology (narrow)0.0010.001
Meta-epidemiology (broad)0.0010.001
Bibliometrics0.0010.001
Science and technology studies0.0010.009
Scholarly communication0.0030.009
Open science0.0020.002
Research integrity0.0030.009
Insufficient payload (model declined to judge)0.0100.004

Machine scores (provisional)

The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.

Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.

Opus teacher head0.034
GPT teacher head0.255
Teacher spread0.222 · how far apart the two teachers sit on this one work
Validation statusscore_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it

Classification

machine, unvalidated

Machine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.

The models applied no category: nothing in the taxonomy fit this work.
Study designTheoretical or conceptual
Domainnot available
GenreMethods

How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".

Quick stats

Citations10
Published2007
Admission routes1
Has abstractyes

Explore more

Same topicUser Authentication and Security SystemsFrench-language works237,207