Skip to content

GCRI · Intelligenza artificiale e sistemi autonomi

Cybersicurezza dell'IA

AI and cybersecurity meet in three ways: AI sistemi can be attacked, AI can help attackers, and AI can help defenders. Models can be poisoned, stolen, manipulated or made to leak data; their supply chains of data, weights and libraries carry new risks. This page sets out how these threats work, how AI systems are secured, and what evidence serves the security teams and regulators responsible.

01Registrare

AI cyber security on the record.

02La domanda

AI cyber security: how the risk develops.

Adversarial methods and model exposure. The dossier below sets out how the risk develops, what is measured, where the data come from, how it is modelled and who decides.

Meccanismo

Models are attack surfaces

Adversarial inputs, prompt injection, data extraction and model theft target the model itself. Classic application vulnerabilities still apply around it.

Meccanismo

The AI supply chain is long

Training data, pretrained weights, libraries and plugins come from many sources. Poisoned data or tampered models can introduce hidden behaviour.

Meccanismo

AI lowers the cost of some attacks

Generative AI can draft phishing, translate lures and help write code, lowering the skill barrier for some attacks. The size of this effect is studied and debated.

Meccanismo

AI assists defence

Machine learning detects anomalies, triages alerts and finds vulnerabilities. Defenders must also secure the AI tools they adopt.

03Simulatore

Una cascata attraverso reti interdipendenti.

A impronta di pericolo elimina gli asset colocalizzati in ogni livello; frammentazione, dipendenza tra livelli e sovraccarico dei sopravvissuti propagano poi il guasto. La vista 3D separa i livelli per profondità. Modello didattico con valori illustrativi dichiarati, non una previsione.

Comandi

Beni per livello
18
Impronta del pericolo
8% della superficie
Beni che dipendono da un altro livello
60%
Capacità di riserva (tolleranza)
30%

Letture

Beni colpiti dal pericolo
7%
Beni ancora funzionanti
50%
Amplificazione del danno iniziale
6.8 times
Passaggi fino al regolamento
3
Numeri, metodi e fonti
PassaggioData center in funzione (%)Reti funzionanti (%)Funzionamento dell'alimentazione (%)
08394100
1727894
2616750
3616722

Metodi

  1. Reti interdipendenti Buldyrev and others (2010), Catastrophic cascade of failures in interdependent networks, Nature 464
  2. Cascate di sovraccarico Motter and Lai (2002), Physical Review E 66
  3. Carico di intermediazione Brandes (2001), Journal of Mathematical Sociology 25
  4. Reti multistrato Kivela and others (2014), Journal of Complex Networks 2
  5. Interdipendenze delle infrastrutture Rinaldi, Peerenboom and Kelly (2001), IEEE Control Systems 21(6)
  6. Ventaglio di incertezza (Monte Carlo sui parametri) Spiegelhalter, Pearson e Short (2011), Visualising uncertainty about the future, Science 333
  7. Sensibilità un fattore alla volta (tornado) Saltelli and others (2008), Global Sensitivity Analysis: The Primer, Wiley

Dati

The model runs on the teaching values set by the controls. To calibrate it for a place, read it against these admitted layers and series, which this page’s map can draw:

  • Strade principali Natural Earth
  • Ferrovie Natural Earth
  • Porti marittimi Natural Earth
  • Aeroporti Natural Earth
  • Estensioni urbane Natural Earth
  • Luci notturne stanotte, interruzioni (banda giorno/notte VIIRS) NASA/NOAA VIIRS (Suomi NPP) · registro nasa.gibs.context-live-layers
  • Eventi naturali (NASA EONET) NASA EONET · registro nasa.eonet.events
Una cascata attraverso reti interdipendenti. Modello didattico illustrativo: funziona con i valori impostati dai controlli, non con dati misurati. Metodi: reti interdipendenti (Buldyrev e altri (2010), Catastrophic cascade of failures in interdependent networks, Nature 464); cascate di sovraccarico (Motter e Lai (2002), Physical Review E 66); carico di intermediazione (Brandes (2001), Journal of Mathematical Sociology 25).
04Tra i domini

AI changes cyber risk in every domain’s control systems.

05Metodi ed evidenze

How AI cyber security is modelled, measured and tested.

Il modello, le sue ipotesi, i suoi dati e la sua incertezza, nella forma in cui uno scienziato o un ingegnere li verificherebbe. Le strutture seguono metodi pubblicati. I valori rappresentati sono illustrativi, non risultati.

Modello di sistema

The couplings between the systems in AI cyber security: signs of influence, a reinforcing and a balancing loop, and cross-links where stress jumps.

++−+Intelligenza artificialeDatiComunitàEconomiaR di rinforzoB bilanciamento
Causal loop diagram from the page’s systems; signs show the direction of influence. Illustrative structure for a system dynamics or network model, not estimated couplings.
Principi guida FAIR · GO FAIR e FORCE11W3C PROV · World Wide Web ConsortiumGuidance on consistent treatment of uncertainties · IPCCCEOS Analysis Ready Data · Comitato sui satelliti per l'osservazione della Terra
06Indicatori

Cosa viene misurato e con quale accuratezza.

Ogni indicatore riporta la sua unità, la sua cadenza, il tipo di fonte su cui si basa e l'etichetta di evidenza che un valore riporterebbe. Non sono mostrati valori: un programma li fornisce sotto la propria autorità.

IndicatoreUnitàCadenzaTipo di fonteEtichetta
Vulnerabilities reported in AI frameworks and librariescount of CVE entriesmensileVulnerability databasesRIPORTATO
Adversarial robustnessaccuracy under a stated attackper evaluationRobustness testingOSSERVATO
Models with signed provenancepercent of deployed modelsper inventoryModel registriesDICHIARATO
AI-related security incidentscount by techniqueannualeIncident reportsRIPORTATO
Time to detect AI system compromisehoursper incidentIncident response recordsOSSERVATO
Organisations with AI security policiespercent of surveyed organisationsper surveyIndustry surveysRIPORTATO
07Fonti

Da dove provengono le evidenze.

Set di dati pubblici e registri istituzionali specifici, ciascuno con la sua classe di licenza. Le fonti soggette ad accordo sono utilizzate solo alle relative condizioni e mai ripubblicate.

FonteCosa fornisceClasse di licenza
MITRE ATLASAdversarial tactics and techniques against AI sistemi, with case studiesPublic, licence to check
NIST adversarial machine learning taxonomyTerminology and taxonomy of attacks and mitigationsPublic domain, US government
National Vulnerability Database and CVE recordsVulnerabilities in AI softwarePublic domain, US government
ENISA reports on AI cybersecurityThreat landscape and good practice for securing AIPublic, ENISA terms
Joint guidance from national cyber agencies on secure AI system developmentSecure design, development, deployment and operationPublic, terms per agency
08Decisioni

Chi decide, e a cosa servono le evidenze.

Studio Intelligenza artificiale e sistemi prepares evidence. The decisions below are taken by the authorities named, under their own rules, and never by the platform.

Chi decideLa domandaEvidenze fornite da questa pagina
Chief information security officersHow to secure AI systems the organisation builds or buysThreat models; supply chain checklists
National cybersecurity agenciesWhat guidance to issue on AI securityIncident and vulnerability evidence
SviluppatoriWhich defences to build into models and applicationsRobustness testing; attack taxonomies

Output prodotti da questo argomento

  • Threat models for AI deployments
  • AI supply chain security checklists
  • Robustness testing protocols
  • Evidence notes on AI-enabled attack trends
09Questioni centrali

What security researchers ask.

  • How can inputs be crafted to make a model misbehave, and how robust are defences?
  • Can an attacker poison training data or tamper with model weights in the supply chain?
  • Can private training data or a model itself be extracted through its outputs?
  • How does AI change attackers’ and defenders’ capabilities?
10Meccanismi

The AI attack surface.

Adversarial examples are inputs altered in ways imperceptible to people that cause models to err. Data poisoning inserts malicious examples into training data to plant backdoors. Membership inference and extraction attacks recover information about training data or replicate a model through repeated queries. For language models, jailbreaks and prompt injection bypass intended restrictions.

On the other side, AI can help attackers find vulnerabilities, write malware variants and craft convincing phishing at scale, and help defenders triage alerts, analyse code and detect anomalies. The balance between the two is an open research question.

11Metodi e dati

Security engineering for AI.

  1. 01

    Red-teaming

    Structured attempts to break a system, combining cybersicurezza e crittografia and artificial intelligence and machine learning expertise. Red-teaming finds problems but cannot prove their absence.

  2. 02

    Threat modelling

    Mapping assets, adversaries and attack paths across data, training, deployment and supply chain.

  3. 03

    Robustness testing

    Measuring performance under adversarial perturbation. Robustness to known attacks does not imply robustness to new ones.

  4. 04

    Supply-chain integrity

    Provenance of datasets, models and dependencies, with signing and verification. Provenance tooling is still maturing.

13Where this page’s references

Where this page’s references lead.

Gli organismi e i documenti alla luce dei quali si legge questa pagina. Stato ed edizioni cambiano; ogni link porta all'organismo emittente.

14Domande

Domande e risposte.

What is data poisoning?
Manipulating training data so that a model learns hidden behaviours or performs worse, for example responding to a trigger in a way the attacker chooses.
What is MITRE ATLAS?
A knowledge base of adversary tactics and techniques against AI systems, modelled on MITRE ATT&CK, with real-world case studies.
Does AI make cyberattacks easier?
It can lower the effort for some tasks, such as writing phishing messages, according to agency assessments. How much it changes overall attack capability is still being studied.
Does the platform publish attack techniques?
Only at a level that supports defence, following responsible disclosure. Details that would mainly enable harm are not published.
How do I report a vulnerability?
Scrivere a il modulo di segnalazione per la sicurezza. La conferma di ricezione fornisce un riferimento e indica il passo successivo.
Is open release of model weights a security risk?
It is a trade-off. Open weights enable scrutiny, research and local control, and also make misuse and the removal of safeguards easier. The balance depends on the model’s capabilities.
Can AI defend against AI-enabled attacks?
It can help, but it is not a complete answer. Basic security practices, monitoring and human expertise remain essential.
RapportiContribuisci

Contribuisci con evidenze da questa pagina.

Rapporti is the Consortium’s quarterly publication. What you know about this subject can become evidence, a case or a correction in the next edition, credited to you and published under CC BY 4.0 where it is your own work.

GCRIIntelligenza artificiale e sistemi autonomiTecnologia esponenzialeEvidenze di Labs
Intelligenza artificiale e sistemi autonomi

Porta un metodo, un set di dati o una domanda.

Informazioni e richieste sono aperte oggi. Una richiesta riceve un riferimento e indica il passo successivo; candidature, pagamenti, assegnazioni di calcolo e registrazioni si aprono solo quando esistono il team, le condizioni e la capacità di revisione.