Methods/Defense

AI vulnerability discovery for defense

Using AI agents to find vulnerabilities in software before attackers do.

38 records1 attack6 capability27 defense4 policy3 findings (1 measured)First recorded 2023-08assistant-drafted

How it works

Agents read code, generate fuzz targets, and reason about crashes; findings are reported to maintainers for fixing.

Known limits

Report volume can outpace maintainers' capacity to triage and patch.

What we know

2 corroborated, 1 qualified

Records over time

RangeLanes
32 of 38 records in view

Use the arrow keys to move between records, Home and End to jump to the first and last, and Enter to select one.

Agents find real bugsAgents in real operationsGated capability, incidents in the labAttackCapabilityDefensePolicyJan 25Jul 25Jan 26Jul 26
Full record · drag to choose a range
202420252026

Select a mark to read the record. Mark size shows editorial significance. Hollow marks are dated to the month. Era bands are editorial labels.

Records in view

32 records · newest first
Nov 2026
Nov 9, 2026
Anthropic releases Fable and Mythos 5.2 with different safeguards and access
CapabilitySystem cardAnthropic

Anthropic released Claude Fable 5.2 for general access and Mythos 5.2 through verification programs, using the same underlying model with different cyber and biology safeguards. Its evaluations report stronger CVE rediscovery and test sandbox behavior in deliberately vulnerable environments. Cross-lab CyScenarioBench scores use different tasks.

Nov 4, 2026
Observational Glasswing study reports more disclosures with limited outcome visibility
DefensePaperQueen Mary University of London

Thomas Mosk’s observational preprint estimates more than doubled public vulnerability disclosures for Glasswing-access firms relative to comparison firms. Measured remediation did not slow among fixed-and-disclosed flaws. A suggestive decline in the exploited share and no statistically detectable count change do not establish fewer attacks; non-random access and limited disclosure visibility constrain the conclusions.

Nov 4, 2026
OpenAI reports Linux discovery, exploit validation and candidate patch pipeline
DefenseEvaluation reportOpenAI

OpenAI describes a Linux research pipeline that produced 2,308 candidate findings, dynamically reproduced 1,724 and developed 948 local privilege-escalation exploits on its stock desktop target. It generated candidate patches for review, rather than reporting that all findings were novel or all fixes had shipped. Hardened-target results are a separate cohort.

Oct 2026
Oct 29, 2026
Epoch AI reports GLM 5.3 exploited the patch gap in six Firefox releases in offline experiments
CapabilityEvaluation reportEpoch AI, Zhipu AI (Z.ai), Mozilla

Epoch AI reports that GLM 5.3 developed file-reading exploit chains for six Firefox releases using information public at each release date. Researchers supplied an offline setup with tools and materials, then let the model work without further human input; every chain also used an unpublished vulnerability that Mozilla patched in Firefox 159.0.

Oct 27, 2026
Anthropic describes five real-software exploit chains developed with an unreleased research model
CapabilityEvaluation reportAnthropic, Google, Apple

Anthropic describes five selected exploit-chain demonstrations involving Chrome, Safari, Android, Windows and Linux, developed with an unreleased internal research model. It reports coordinated vendor fixes and little human involvement after initial prompting, with different outcomes including code execution, data access and denial of service.

Oct 21, 2026
OpenAI describes coordinating PHP's disclosure of two critical flaws AI review brought to light, and proposes a response framework
DefenseVulnerability disclosureOpenAI, Wiz, Cloudflare

OpenAI describes coordinating fixes for two critical PHP flaws after Calif identified one using AI and Wiz independently found both; PHP released fixes on October 21, and OpenAI proposed a disclosure framework. Follow-up evidence includes CISA’s October 26 listing of both flaws as known exploited and Palo Alto Networks’ report of active exploitation against internet-exposed management interfaces. Wiz’s October 27 report describes declining observed exposure and failed chains in its own captured attempts; this does not establish an absence of exploitation elsewhere or a global compromise count.

Oct 20, 2026
24 model setups: an engineered prompt with code-graph context lifts smaller models' vulnerability finding, not a frontier model's
CapabilityPaperUniversity of Toronto

Researchers at Georgian and the University of Toronto ran Vercel Labs' open-source DeepSec harness with 24 model–runner configurations on a production TypeScript codebase whose vulnerabilities they say were non-public at scan time. No configuration covered all 265 confirmed security root causes, and moving from a minimal prompt to an engineered prompt with CodeQL-derived call-graph context raised criticality-weighted yield by 20–86% for three smaller or cheaper models but by 2% for GPT-5.5. The paper is published in the 2026 CAMLIS proceedings (PMLR 349).

Oct 12, 2026
Collinear's CWE-bench: 120 private audit-and-repair tasks, with three of eleven agents tied at 68% pass@1
DefenseBenchmarkCollinear

Collinear researchers introduce CWE-bench, a private benchmark of 120 repository audit-and-repair tasks built on real open-source repositories with injected vulnerabilities across 73 CWE types, eight languages and all ten OWASP Top 10:2025 categories; agents receive a threat model but not the weakness, its location or the number of vulnerable sites. A patch passes only if a programmatic verifier finds the exploits blocked and existing functionality tests passing, and the authors check the verifier against a three-judge agentic panel. They report that Grok 4.7, Gemini 4 Argon and GPT-6 Astra tie at 68% pass@1, that Claude Opus 5.5 reaches 67%, and that pass@4 for the tied agents ranges from 74% to 81%; the results are author-run on a task set outsiders cannot inspect.

Oct 8, 2026
Anthropic launches critical-infrastructure support and free open-source scanning
DefenseProgramAnthropic

Anthropic launches its Cyber Mission with a Critical Infrastructure Defense Program and an opt-in OSS Scanner offering free periodic security scans to open-source maintainers. Scanner reports are model-generated and sent without human review; Anthropic continues human-verified disclosure for projects unable to absorb that volume. The announcement describes efforts and expected performance rather than an independently measured reduction in security risk.

Oct 7, 2026
Microsoft FORGE reports scaled vulnerability discovery and validation bottlenecks
DefenseEvaluation reportMicrosoft

Microsoft’s FORGE reports distinguish internally validated vulnerability findings from maintainer acceptance and shipped fixes. A November 11 Linux follow-up traces 26,502 raw findings to 621 PoC-backed cases, 59 upstream reports, 33 assigned CVEs and 38 reports with fixes merged into mainline. Expert triage, deduplication and maintainer review took months; these cohort-specific figures should not be added to earlier program totals.

Sep 2026
Sep 30, 2026
Google's threat intelligence group reports CVE disclosures doubled in 2026 and profiles likely AI-discovered vulnerabilities
AttackMisuse reportGoogle Threat Intelligence Group, Hacktron AI

Google Threat Intelligence Group (GTIG) reports that monthly CVE disclosures rose from 5,045 in January 2026 to 10,477 in July and 10,740 in August, and that exploited vulnerabilities averaged 18 per month from January to August 2026 against 10.5 per month in 2025. GTIG says zero-day exploitation rose only marginally and suggests rapid weaponization of n-days drove most of the growth, and that vulnerabilities it identified as likely AI-discovered skew toward Medium risk ratings and remote code execution. It also counts 2,076 cumulative AI-related CVEs (January 2025 to August 2026) and reports in-the-wild exploitation of a handful, including a flaw in LiteLLM's MCP server preview endpoints.

Sep 30, 2026
Google DeepMind says Gemini 4 Argon is rolling out to trusted cyber defenders and will ship to them without cyber guardrails
CapabilityAccess programGoogle DeepMind, Google, Wiz

Google DeepMind announced Gemini 4 Argon on 2026-09-30 and says it is rolling out to a set of trusted cyber defenders through the Fairwind Program, and says it will release the model without cyber guardrails for those defenders and Google's internal teams. Google claims the model can autonomously find, validate and patch critical vulnerabilities, ties for first on CWE-bench v1 at 68%, and improves on Gemini 3.8 Flash Cyber in vulnerability discovery, without giving numbers for the comparison. It says broad release awaits further safeguard work in four areas (misuse, prompt injection, misalignment and hardening), citing its Frontier Safety Framework for the misuse safeguards; all of these are Google-reported claims that have not been independently verified.

Sep 28, 2026
RedHerring decoys cut real vulnerabilities found by five open-weight agents by 38.7% to 60.4% at matched budgets
DefensePaperHong Kong University of Science and Technology

HKUST researchers argue that agentic vulnerability discovery forms many hypotheses but can verify only some within a fixed budget, so verification effort is something a defender can steer. Their RedHerring system inserts decoy code paths that look like real CVE-style bugs but are unreachable behind a predicate the defender can certify with private information. In 70 OSS-Fuzz instances, the authors report that five open-weight models run in Claude Code confirmed 38.7% to 60.4% fewer real crashes, with 30.6% to 51.5% of completion tokens spent on decoys.

Sep 28, 2026
Artificial Analysis launches a Cyber Index and industry alliance scoring models on source-level vulnerability finding and fixing
DefenseBenchmarkArtificial Analysis, Collinear, IBM

Artificial Analysis announced a Cyber Index Alliance, with Collinear AI, IBM, NVIDIA and Vercel as launch partners, and a v1 Cyber Index that combines three source-code evaluations: repository audit and patch (CWE-Bench-AA, 120 private tasks), finding expert-verified vulnerabilities (DeepsecBench-AA) and discovering, reproducing and patching memory-safety bugs (CyberGym-E2E-AA, 131 tasks). Artificial Analysis says it excludes exploit building, runs all three on its open-source Stirrup harness, and reports safety refusals separately from scores. The article reports failure-mode statistics and says several frontier models refuse almost every CyberGym-E2E-AA task; per-model index scores appear only in charts.

Sep 24, 2026
Google's PageBreak agent finds over 500 XSS bugs in its own web apps using deterministic validators
DefenseTool releaseGoogle

Google's Product Security team describes PageBreak, an internal agent mostly using Gemini models that hunts vulnerabilities in Google's first-party web applications and only reports findings confirmed by non-AI validators against running applications. Google reports over 500 XSS vulnerabilities found with near-zero false positives, while apps on its high-assurance web frameworks yielded only 2 XSS bugs as of 4 September 2026.

Sep 23, 2026
OpenAI extends its Daybreak cyber-defense access to the Government of Ukraine for civilian infrastructure
CapabilityAccess programOpenAI, ENISA, Ukraine Ministry of Digital Transformation

OpenAI announced on 2026-09-23 that it will give the Government of Ukraine access to its Daybreak program, working with Ukraine's Ministry of Digital Transformation, so Ukrainian teams can find software vulnerabilities and develop and test fixes for civilian infrastructure. OpenAI describes Daybreak as giving defenders access to advanced AI for authorized security work, and says it has provided its cyber models to defenders in France, Germany, Poland and elsewhere in Europe. It reports that ENISA used the models to identify vulnerabilities in software used across EU institutions and that CERT Polska used them to help find six router-software vulnerabilities; these outcomes are OpenAI's account.

Jun 2026
Jun 12, 2026
US export-control directive forces Anthropic to suspend Fable 5 and Mythos 5 over safeguard bypass
PolicyRegulationUS Department of Commerce, Anthropic

Anthropic said the US government issued an export control directive, citing national security authorities, barring access to Fable 5 and Mythos 5 by foreign nationals, after officials said they had found a way to jailbreak Fable 5's safeguards. Anthropic said the net effect was that it had to disable both models for all customers to comply, while other Claude models stayed available. Anthropic disputed the rationale, arguing the demonstrated vulnerabilities were minor and that the standard applied industry-wide would halt new frontier deployments.

May 2026
May 22, 2026
Glasswing discovery and disclosure updates distinguish reported bugs from known upstream patches
DefenseProgramAnthropic, Cloudflare, Mozilla

Anthropic’s May Glasswing update reports high- and critical-severity discovery outpacing triage and patching. Its November 2 disclosure dashboard records 5,087 reports across 680 projects and 629 known upstream patches across all severities and multiple Claude models. Different cohorts, triage routes and incomplete patch visibility prevent treating these snapshots as a comparable patch-rate trend.

May 11, 2026
UK NCSC issues ten questions for organizations using AI models to find vulnerabilities
PolicyGuidanceUK National Cyber Security Centre

The head of the NCSC's Vulnerability Management Group published ten questions for organizations considering AI-driven vulnerability discovery. The questions stress having a process to triage and fix findings, prioritizing exploitable issues, weighing data, permission, legal and jurisdiction risks of the chosen model, starting with the external attack surface, and planning for future models.

Apr 2026
Apr 7, 2026
Anthropic launches Project Glasswing to give defenders early access to Claude Mythos Preview
DefenseProgramAnthropic, Amazon Web Services, Apple

Anthropic launched Project Glasswing with AWS, Apple, Broadcom, Cisco, CrowdStrike, Google, JPMorganChase, the Linux Foundation, Microsoft, NVIDIA and Palo Alto Networks to use the unreleased Claude Mythos Preview for defensive security work, extending access to over 40 more organizations that maintain critical software. Anthropic committed up to $100M in usage credits and $4M in donations to open-source security groups, and reports Mythos Preview found thousands of high-severity vulnerabilities, including in every major operating system and browser.

Mar 2026
Mar 9, 2026
OSS-CRS makes AIxCC reasoning systems runnable locally; OpenSSF adopts it as a sandbox project
DefenseTool releaseGeorgia Institute of Technology, Microsoft, Team Atlanta

Researchers led by Georgia Tech released OSS-CRS, a locally deployable framework for running and combining AIxCC cyber reasoning systems, noting that all seven open-sourced finalist systems depended on competition cloud infrastructure that no longer exists. Porting the winning Atlantis system, they found 10 previously unknown bugs (three high severity) in 8 OSS-Fuzz projects; OpenSSF welcomed OSS-CRS into its AI/ML Security Working Group in April 2026.

Feb 2026
Feb 20, 2026
Anthropic releases Claude Code Security in limited preview to scan code and propose patches
DefenseTool releaseAnthropic

Anthropic released Claude Code Security as a limited research preview for Enterprise and Team customers, with expedited free access for open-source maintainers. The tool reasons about data flow across a codebase, re-examines each finding in a multi-stage verification pass, assigns severity and confidence ratings, and proposes patches that are applied only with human approval.

Feb 7, 2026
AIxCC SoK finds stability decided results and many validated AI patches were still semantically wrong
DefensePaperGeorgia Institute of Technology, Texas A&M University, DARPA

A systematization-of-knowledge paper by organizers and competitors analyzes AIxCC's design, the seven finalist architectures and results beyond the scoreboard. It reports that system stability and accuracy penalties decided rankings, that LLM-based systems found vulnerabilities a fuzzing baseline missed, and that among patches passing all automatic validation, manual review found semantic errors in 38-46% from baseline agents; the top two systems had 83.8% and 79.2% competition-scored patch accuracy.

Feb 5, 2026
Anthropic reports over 500 human-validated high-severity open-source vulnerabilities found with Claude Opus 4.6
DefensePaperAnthropic, Ghostscript, OpenSC

Anthropic's Frontier Red Team reports that Claude Opus 4.6, run in a VM with standard tools but no custom harness, found and validated more than 500 high-severity vulnerabilities in open-source software, focusing on memory corruption that can be confirmed with sanitizers. Every bug was validated before reporting, initially by Anthropic researchers who also wrote patches and later with external researchers; examples include Ghostscript, OpenSC and CGIF.

Jan 2026
Jan 9, 2026
Anthropic's next-generation Constitutional Classifiers cut overhead to about 1% using probe cascades
DefensePaperAnthropic

Anthropic describes Constitutional Classifiers++, a cascade in which a cheap linear probe on model activations screens all traffic and escalates flagged exchanges to a probe-classifier ensemble. It reports roughly 1% added compute if applied to Claude Opus 4.0 traffic (the first generation added 23.7%) and a 0.05% refusal rate on harmless queries over one month of Claude Sonnet 4.5 traffic. Red-teamers found no universal jailbreak in over 1,700 hours.

Oct 2025
Oct 30, 2025
OpenAI announces Aardvark, a GPT-5 agent that finds, validates and proposes patches for vulnerabilities
DefenseTool releaseOpenAI

OpenAI announced Aardvark, a GPT-5-powered agent in private beta that builds a threat model of a repository, scans commits, tries to trigger suspected flaws in a sandbox, and attaches Codex-generated patches for human review. OpenAI reports 92% recall on known and synthetically introduced vulnerabilities in its 'golden' repositories and ten CVEs from open-source scanning, and planned pro-bono scanning for some non-commercial projects.

Oct 3, 2025
Anthropic says it trained Claude Sonnet 4.5 for defensive vulnerability finding and patching
DefensePaperAnthropic, HackerOne, CrowdStrike

Anthropic reports that a small team focused Claude Sonnet 4.5 training on finding and patching vulnerabilities and on testing simulated security infrastructure, while avoiding enhancements that clearly favour offence. It reports Sonnet 4.5 results on Cybench and CyberGym, a preliminary patching study in which 15% of patches were judged semantically equivalent to human references, and invites work on SOC and SIEM automation.

Aug 2025
Aug 8, 2025
AIxCC final: Team Atlanta wins as systems patch 43 of 54 found synthetic bugs and find 18 real ones
DefenseCompetitionDARPA, ARPA-H, Team Atlanta

DARPA reports that seven finalist cyber reasoning systems analyzed over 54 million lines of code, found 54 unique synthetic vulnerabilities in 63 challenges and patched 43, and found 18 real non-synthetic vulnerabilities with 11 patches. Team Atlanta won $4 million, Trail of Bits $3 million and Theori $1.5 million; DARPA and ARPA-H added $1.4 million for real-world integration and four systems were open-sourced on the day.

Aug 4, 2025
Google reports Big Sleep found and reproduced 20 vulnerabilities in open-source projects
DefenseVulnerability disclosureGoogle, Google DeepMind, Google Project Zero

Google's vice president of security announced that Big Sleep had reported 20 vulnerabilities, mostly in open-source projects such as FFmpeg and ImageMagick, with details withheld pending fixes. A Google spokesperson told TechCrunch each flaw was found and reproduced by the agent without human intervention, with a human expert reviewing reports before submission.

Jul 2025
Jul 15, 2025
Google says Big Sleep found SQLite CVE-2025-6965 before attackers could exploit it
DefenseVulnerability disclosureGoogle, Google DeepMind, Google Project Zero

Google reports that, working from Google Threat Intelligence information, the Big Sleep agent found a critical SQLite memory-corruption flaw (CVE-2025-6965) that Google says was known only to threat actors and at risk of exploitation. Google says it reported the flaw for patching before attackers could exploit it, says it believes this is the first time an AI agent directly foiled an in-the-wild exploitation effort, and says Big Sleep is being applied to open-source projects.

Jul 10, 2025
EU GPAI Code of Practice Safety and Security chapter lists cyber offence as a specified systemic risk
PolicyFrameworkEuropean Commission

The European Commission received the final General-Purpose AI Code of Practice, whose Safety and Security chapter applies to providers of models with systemic risk under Article 55 of the AI Act. The chapter treats cyber offence as one of four specified systemic risks, requires a security goal covering non-state external and insider threats, and sets serious incident reporting deadlines that include five days for serious cybersecurity breaches.

All records