Does Frontier AI Strengthen Cyber Defense? Evidence from Project Glasswing
Organizations: Queen Mary University of London. Mile End Road London E1 4NS
Abstract
Defenders and attackers race to find the same software flaws. I use the launch of Project Glasswing, which gave a small group of large US tech firms early access to Claude Mythos Preview, a model for finding software vulnerabilities, to study how access to frontier AI affects defenders, external researchers, and attackers. Using the universe of publicly disclosed vulnerabilities, I find that firms with access more than doubled their disclosures relative to Chinese and Russian tech firms without access, disclosed about 150 percent more high-impact vulnerabilities, and fixed reported flaws as quickly as before. As treated firms found more flaws themselves, external researchers, who sell their findings through bug bounties, reported about half as many vulnerabilities in treated firms' code and more in other firms' code, without leaving the market. There is no evidence that attackers exploited the additional vulnerabilities: the share exploited within a month of disclosure fell by about two thirds, mainly because a smaller share was already exploited when disclosed, while the number exploited did not change. Frontier AI helped defenders find and fix more flaws before attackers exploited them, but its restricted release creates an asymmetry in software security between firms in different countries.
Figures & tables
| Panel A: vulnerabilities by group of firms | ||||||
|---|---|---|---|---|---|---|
| Full sample | Pre-treatment (Jan 2024–Mar 2026) | |||||
| Glasswing | Other non- | Chinese | ||||
| Obs. | Mean | SD | firms | Chinese firms | firms | |
| Observations (pre-treatment) | 9,246 | 90,077 | 4,577 | |||
| Glasswing firms | 177,104 | 0.121 | 0.326 | |||
| Other non-Chinese firms | 177,104 | 0.845 | 0.361 | |||
| Comparison group: Chinese firms | Alternative comparison groups | |||||||
| (1) | (2) | (3) | (4) | (5) | (6) | (7) | (8) | |
| Baseline | Poisson | Excluding Anthropic | Glasswing trend | Placebo April 2025 | Non-Chinese firms | Chinese and Russian firms | All firms | |
| Glasswing Post | 0.784*** | 1.136*** | 0.862*** | 0.687*** | -0.034 | 0.339* | 0.785*** | 0.778*** |
| (0.206) | (0.294) | (0.206) | (0.208) | (0.102) | (0.199) | (0.206) | (0.206) | |
| Glasswing trend (months) | 0.0184*** | |||||||
| (0.0069) | ||||||||
| (1) | (2) | (3) | (4) | (5) | (6) | (7) | (8) | |
| All | AI-exposed | Not AI-exposed | Complex weakness | Machine- verifiable | High impact | Remote, no credentials | Many releases | |
| Glasswing Post | 0.784*** | 0.642*** | 0.768*** | 0.720*** | 0.825*** | 0.919*** | 0.747*** | 0.580*** |
| (0.206) | (0.224) | (0.226) | (0.226) | (0.195) | (0.209) | (0.187) | (0.182) | |
| Pre-launch mean of , Glasswing firms | 29.165 | 8.405 | 20.038 | 15.506 | 12.937 | 14.468 | 4.949 | 19.076 |
| Pre-launch mean of , Chinese firms | 0.589 | 0.267 | 0.321 | 0.147 | 0.441 | 0.327 | 0.124 | 0.089 |
| Firm FE | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes |
| A. Vendor fix of flaws reported by the Zero Day Initiative | B. Fix by Linux distributions after publication | ||||
| (1) | (2) | (3) | (4) | (5) | |
| days report to fix | Fixed within 90 days of report (%) | Fixed within 30 days (%) | Fixed within 90 days (%) | days if fixed | |
| Glasswing Post | -0.329 | 13.394 | -3.123 | -4.495 | 0.340 |
| (0.232) | (21.348) | (9.155) | (3.510) | (0.320) | |
| Pre-access mean, Glasswing vulnerabilities | 4.702 | 28.070 | 78.857 | 83.429 | 1.267 |
| Pre-access mean, other vulnerabilities | 4.589 | 35.106 | 61.004 | 74.884 | 2.335 |
| Credited vulnerabilities ( ) in the month | Extensive margin | ||||||
| (1) | (2) | (3) | (4) | (5) | (6) | (7) | |
| All | Glasswing code | Other non-Chinese code | Chinese code | AI-exposed | Not AI-exposed | Not active | |
| Post | 0.489* | -0.832** | 0.891** | 2.266** | 0.463* | 0.289 | 0.003 |
| (0.273) | (0.387) | (0.398) | (1.068) | (0.249) | (0.270) | (0.008) | |
| Expansion | 0.084 | 0.204 | 0.141 | 3.559** | -0.083 | 0.196 | 0.000 |
| (0.289) | (0.384) | (0.338) | (1.695) | (0.368) | (0.319) | (0.007) | |
| Glasswing | Large non-Chinese | Chinese | All other | |
| firms | firms (26) | firms | firms | |
| Vulnerabilities | 9,246 | 5,287 | 3,906 | 84,790 |
| Exploited by October 28, 2026 (%) | 2.39 | 1.42 | 1.61 | 1.49 |
| Exploited within 90 days of publication (%) | 1.83 | 1.10 | 0.61 | 0.98 |
| Public exploit code by October 28, 2026 (%) | 4.61 | 2.53 | 2.25 | 4.75 |
| Exploited vulnerabilities | 221 | 75 | 63 | 1,265 |
| (1) | (2) | (3) | (4) | (5) | (6) | (7) | |
| Exploited 1 month | Exploited 3 months | Zero-day | Exploit code 1 month | Exploit code 3 months | EPSS 1 month | EPSS 3 months | |
| Panel A: Chinese firms as comparison group | |||||||
| Glasswing Post | -0.921** | -0.783 | -0.848** | -1.527*** | -0.711 | -0.100 | 0.136 |
| (0.435) | (0.529) | (0.361) | (0.527) | (0.589) | (0.132) | (0.161) | |
| Pre-launch mean, Glasswing firms | 1.419 | 1.729 | 1.197 | 2.838 | 3.636 | 0.160 | 0.298 |
| Pre-launch mean, Chinese firms | 0.575 | 0.806 | 0.575 | 1.381 | 1.611 | 0.221 | 0.476 |
Appendix figures & tables13 assets
Supplementary material from the paper’s appendix.
Appendix
| Source | Vulnerabilities covered | With acknowledgment (%) |
|---|---|---|
| CVE List (credits field) | 176,384 | 43.4 |
| Microsoft Security Update Guide | 21,615 | 30.3 |
| Chrome release notes | 3,324 | 100.0 |
| Cisco security advisories | 987 | 99.7 |
| Apple security notes | 2,412 | 96.9 |
| Android security bulletins | 934 | 100.0 |
| Vulnerabilities of treated and control firms | |||||||
|---|---|---|---|---|---|---|---|
| Full sample | Pre-treatment (Jan 2024–Mar 2026) | ||||||
| Obs. | Mean | SD | Treated firms | Non-treated firms | Difference | -stat. | |
| Found by the vendor’s internal team: explicit only | 176,384 | 0.040 | 0.195 | 0.133 | 0.014 | 0.118*** | 33.37 |
| Found by the vendor’s internal team: including inferred | 176,384 | 0.147 | 0.354 | 0.247 | 0.094 | 0.153*** | 33.41 |
| Finder unknown: explicit only | 176,384 | 0.270 | 0.444 | 0.251 | 0.259 | -0.008* | -1.65 |
| Finder unknown: including inferred | 176,384 | 0.163 | 0.369 | 0.137 | 0.180 | -0.043*** | -11.22 |
| All vulnerabilities | Treated firms | Control firms | ||
|---|---|---|---|---|
| Weakness type | Number | Share (%) | Share (%) | Share (%) |
| AI-exposed: memory safety | 27,841 | 17.3 | 36.3 | 19.8 |
| Out-of-bounds write and buffer overflow | 12,156 | 7.6 | 13.9 | 8.4 |
| Out-of-bounds read | 4,383 | 2.7 | 6.7 | 4.8 |
| Use after free, double free, invalid free | 5,348 | 3.3 | 11.6 | 2.9 |
| Null pointer dereference | 3,386 | 2.1 | 0.9 | 1.1 |
| Vulnerabilities per month | |||||
| Group | Firm | Pre | Launch | Expansion | Total |
| Glasswing | Microsoft | 99.6 | 177.5 | 592.2 | 5,414 |
| Glasswing | 70.0 | 267.0 | 671.0 | 5,108 | |
| Glasswing | IBM | 47.7 | 49.0 | 243.8 | 2,362 |
| Glasswing | Apple | 56.3 | 59.0 | 128.8 | 2,154 |
| Glasswing | Cisco | 27.8 | 29.5 | 74.8 | 1,109 |
| 2024 | Oct 2025 to | Apr 2026 to | |
| Mar 2026 | Sep 2026 | ||
| A. Chinese vulnerabilities per month, by disclosure channel | |||
| All Chinese vulnerabilities | 180.1 | 157.2 | 192.7 |
| Identifier issued by the Chinese vendor | 38.7 | 45.7 | 61.5 |
| Identifier issued by an independent issuer | 108.8 | 99.2 | 123.2 |
| of which by VulDB | 37.6 | 70.2 | 74.2 |
| Glasswing firm left out | Glasswing Post | Pre-access per month | Firm-months | Clusters |
|---|---|---|---|---|
| None (Table 2 , column 1) | 0.784*** (0.206) | 3,360 | 280 | |
| Microsoft | 0.740*** (0.218) | 101.3 | 3,348 | 279 |
| 0.691*** (0.198) | 84.8 | 3,348 | 279 | |
| Apple | 0.802*** (0.222) | 61.3 | 3,348 | 279 |
| IBM | 0.757*** (0.221) | 47.1 | 3,348 | 279 |
| Cisco | 0.789*** (0.223) | 25.8 | 3,348 | 279 |
| Method | Estimate | s.e. | -value | 95% confidence set |
|---|---|---|---|---|
| Firm-clustered (CR1), as in the paper | 0.784 | 0.206 | 0.0001 | [0.38, 1.19] |
| CR2 with Bell-McCaffrey degrees of freedom (13.2) | 0.784 | 0.204 | 0.0020 | [0.34, 1.22] |
| Wild cluster bootstrap, null imposed (Rademacher weights) | 0.784 | 0.0037 | [0.34, 1.24] | |
| Wild cluster bootstrap, null imposed (Webb weights) | 0.784 | 0.0023 | [0.34, 1.20] | |
| Ibragimov and Müller: t-test on 13 firm-by-firm estimates | 0.785 | 0.204 | 0.0023 | [0.34, 1.23] |
| Residual bootstrap without heteroskedasticity correction (Conley and Taber type) | 0.784 | 0.0000 | [0.64, 0.92] |
| (1) | (2) | (3) | (4) | |
| Baseline | Poisson | Excluding Anthropic | Glasswing trend | |
| Glasswing Post (Mythos Preview) | 0.486** | 1.035*** | 0.589*** | 0.438* |
| (0.235) | (0.343) | (0.228) | (0.238) | |
| Glasswing Post Opus 4.6 | 0.452*** | 0.179 | 0.415** | 0.373** |
| (0.161) | (0.240) | (0.169) | (0.147) | |
| Glasswing trend (months) | 0.0126* |
| Firm-month, | Vulnerability level | |||||
| (1) | (2) | (3) | (4) | (5) | (6) | |
| : gap 7 days | : gap 8–90 days | : gap 90 days | : no CVE identifier | Gap 7 days (%) | ||
| Glasswing Post | 0.524** | 0.941*** | 0.063 | -0.004 | 27.745** | -1.422*** |
| (0.257) | (0.298) | (0.180) | (0.005) | (11.914) | (0.483) | |
| Pre-launch mean, Glasswing firms | 3.291 | 11.861 | 14.013 | 0.000 | 11.285 | 4.114 |
| Pre-launch mean, Chinese firms | 0.342 | 0.156 | 0.044 | 0.046 | 63.061 | 1.801 |
| External | Including inferred | |||||||
| (1) | (2) | (3) | (4) | (5) | (6) | (7) | (8) | |
| All | Internal team | All external | Other organization | Individual | Unknown | Internal | Unknown | |
| Glasswing Post | 0.784*** | 0.576** | 0.650*** | 0.587*** | 0.559*** | 0.580*** | 0.603** | 0.464** |
| (0.206) | (0.272) | (0.161) | (0.157) | (0.143) | (0.213) | (0.264) | (0.209) | |
| Pre-launch mean of , Glasswing firms | 29.165 | 3.949 | 17.899 | 7.038 | 10.861 | 7.316 | 7.582 | 3.684 |
| Post-launch mean of , Glasswing firms | 115.532 | 31.870 | 54.000 | 21.195 | 32.805 | 29.662 | 43.026 | 18.506 |
| (1) | (2) | (3) | (4) | (5) | (6) | |
| Exploited 1 month | Exploited 2 months | Exploited 3 months | Zero-day | N-day 3 months | Exploit code 3 months | |
| Glasswing Post | 0.100 | -0.225 | -0.413 | -0.347 | -0.225 | 0.503* |
| (0.445) | (0.363) | (0.375) | (0.488) | (0.659) | (0.266) | |
| Firm and month FE | Yes | Yes | Yes | Yes | Yes | Yes |
| Mean of dependent variable | 0.198 | 0.220 | 0.226 | 0.157 | 0.069 | 0.403 |
| Observations | 493 | 522 | 522 | 406 | 464 | 667 |
Cybersecurity
Cybersecurity develops methods to protect digital and cyber-physical systems from attacks, vulnerabilities, data loss, and misuse.
See all 379 papers