Towards Cybersecurity Superintelligence: from AI-guided humans to human-guided AI
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866918328487903232 |
|---|---|
| author | Mayoral-Vilches, Víctor Rass, Stefan Pinzger, Martin Gil-Uriarte, Endika Ayucar-Carbajo, Unai Ruiz-Alcalde, Jon Ander de Torres, Maite del Mundo Sanz-Gómez, María Balassone, Francesco Veas-Chavez, Cristóbal R. J. Turiel, Vanesa Glera-Picón, Alfonso Sánchez-Prieto, Daniel Salvatierra, Yuri Zabalegui-Landa, Paul Cabrera-Álvarez, Ruffino Reydel Mayoral-Pizarroso, Patxi |
| author_facet | Mayoral-Vilches, Víctor Rass, Stefan Pinzger, Martin Gil-Uriarte, Endika Ayucar-Carbajo, Unai Ruiz-Alcalde, Jon Ander de Torres, Maite del Mundo Sanz-Gómez, María Balassone, Francesco Veas-Chavez, Cristóbal R. J. Turiel, Vanesa Glera-Picón, Alfonso Sánchez-Prieto, Daniel Salvatierra, Yuri Zabalegui-Landa, Paul Cabrera-Álvarez, Ruffino Reydel Mayoral-Pizarroso, Patxi |
| contents | Cybersecurity superintelligence -- artificial intelligence exceeding the best human capability in both speed and strategic reasoning -- represents the next frontier in security. This paper documents the emergence of such capability through three major contributions that have pioneered the field of AI Security. First, PentestGPT (2023) established LLM-guided penetration testing, achieving 228.6% improvement over baseline models through an architecture that externalizes security expertise into natural language guidance. Second, Cybersecurity AI (CAI, 2025) demonstrated automated expert-level performance, operating 3,600x faster than humans while reducing costs 156-fold, validated through #1 rankings at international competitions including the $50,000 Neurogrid CTF prize. Third, Generative Cut-the-Rope (G-CTR, 2026) introduces a neurosymbolic architecture embedding game-theoretic reasoning into LLM-based agents: symbolic equilibrium computation augments neural inference, doubling success rates while reducing behavioral variance 5.2x and achieving 2:1 advantage over non-strategic AI in Attack & Defense scenarios. Together, these advances establish a clear progression from AI-guided humans to human-guided game-theoretic cybersecurity superintelligence. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2601_14614 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Towards Cybersecurity Superintelligence: from AI-guided humans to human-guided AI Mayoral-Vilches, Víctor Rass, Stefan Pinzger, Martin Gil-Uriarte, Endika Ayucar-Carbajo, Unai Ruiz-Alcalde, Jon Ander de Torres, Maite del Mundo Sanz-Gómez, María Balassone, Francesco Veas-Chavez, Cristóbal R. J. Turiel, Vanesa Glera-Picón, Alfonso Sánchez-Prieto, Daniel Salvatierra, Yuri Zabalegui-Landa, Paul Cabrera-Álvarez, Ruffino Reydel Mayoral-Pizarroso, Patxi Cryptography and Security Cybersecurity superintelligence -- artificial intelligence exceeding the best human capability in both speed and strategic reasoning -- represents the next frontier in security. This paper documents the emergence of such capability through three major contributions that have pioneered the field of AI Security. First, PentestGPT (2023) established LLM-guided penetration testing, achieving 228.6% improvement over baseline models through an architecture that externalizes security expertise into natural language guidance. Second, Cybersecurity AI (CAI, 2025) demonstrated automated expert-level performance, operating 3,600x faster than humans while reducing costs 156-fold, validated through #1 rankings at international competitions including the $50,000 Neurogrid CTF prize. Third, Generative Cut-the-Rope (G-CTR, 2026) introduces a neurosymbolic architecture embedding game-theoretic reasoning into LLM-based agents: symbolic equilibrium computation augments neural inference, doubling success rates while reducing behavioral variance 5.2x and achieving 2:1 advantage over non-strategic AI in Attack & Defense scenarios. Together, these advances establish a clear progression from AI-guided humans to human-guided game-theoretic cybersecurity superintelligence. |
| title | Towards Cybersecurity Superintelligence: from AI-guided humans to human-guided AI |
| topic | Cryptography and Security |
| url | https://arxiv.org/abs/2601.14614 |