Narrative · ↗ rising AI is the defenders' bug hunter All narratives · 10 supporting stories · 1 from Hungary and the region
„Frontier models reportedly find thousands of zero-day vulnerabilities; if defenders receive them first, AI favors defense.”
Counter-narrative The same capability enables attackers to produce exploits at an industrial scale, while mass flaw discovery causes outages. This is illustrated by governments and companies patching all night due to the PHP flaw.
Voices Anthropic (Project Glasswing), Mozilla, OpenAI, Google Project Zero, US official, AP
Analysis According to the narrative, the models' bug-finding capability tips the balance toward defenders if access is rationed in a controlled way. It is fed by lab announcements, and the narrative also justifies keeping the models closed. AI labs have an interest in presenting their capabilities as defensive tools. Its weak point is patching capacity. According to Anthropic, discovery has accelerated so much that disclosure deadlines need to be shortened, and the PHP case showed the burden of patching. Google already writes that attackers are using AI agents to turn patches into exploits.
Evidence: moderate — The 22 security bugs fixed by Mozilla and the PHP outages are measured facts. The “thousands of zero-days”, however, is the developer's own statement, the claim about classified systems comes from an unnamed official, and the 948 exploits are a figure from a single statement.
What's at stake: The stake is whether AI bug hunting increases defenders' advantage or, by overloading repair capacity, attackers'.
Narrative sub-threads The narrative splits into 4 branches; a story may belong to several branches, and some may not appear in the main narrative's list.
→ lasting · evidence: moderate bug-hunting results from frontier models Developers, their partners and researchers say AI models and agents are finding large numbers of severe bugs in real-world software. Examples include OpenSSL, Firefox, classified US government systems, Android and Java. An OpenAI AI pipeline generated 948 Linux privilege-escalation exploits in 30 hours.
This branch rests on the fact of discovery: it reports manufacturer and partner figures (500+, 22, 271, 948) and does not address the distribution of access or the disclosure system.
11 stories · 0 of them Hungarian and regional → lasting · evidence: strong Early access for vetted defenders AI labs and states release cyber-capable models through controlled programs. Government bodies, EU institutions, open-source maintainers and a narrow circle of companies receive them first. The aim is to secure an advantage for defenders.
Here the mechanism is distribution and gatekeeping: Trusted Access, Daybreak, Gold Eagle, ENISA–Mythos. The actors are governments and agencies, not the vulnerabilities themselves.
11 stories · 1 of them Hungarian and regional ↗ rising · evidence: strong Overload of the disclosure and patching process The vulnerability management system is under strain. CVE submissions have increased by 263 percent since 2020, prompting NIST to switch to risk-based NVD processing. AI-slop reports are contaminating the CVE process with false vulnerabilities. Citing AI-accelerated discovery, Anthropic and Google Project Zero are shortening the disclosure deadline to 30 days by 2028. A PHP vulnerability discovered by AI caused government websites in several countries to go offline, including magyarorszag.hu.
This branch shows the institutional consequence: NIST, CVE Root, Project Zero, and Anthropic deadlines, plus an actual shutdown. The actors are owners of vulnerability management systems and operators.
8 stories · 1 of them Hungarian and regional ↗ rising · evidence: moderate Who will get it first: an attacker race That defenders gain the upper hand is not guaranteed. Cybercriminal groups and Chinese hackers also use AI to find zero-day bugs, attackers use AI agents to turn patches into exploits, and the cost of zero-day hunting could drop to a few dollars. However, according to a VulnCheck researcher, only 1% of AI-discovered vulnerabilities have so far been exploited in live attacks.
This branch examines the narrative's premise from the attacker's side: it relies on Google, Amazon, and Anthropic observations, as well as VulnCheck data, not on the labs' own defensive results.
7 stories · 0 of them Hungarian and regional Sub-thread stories bug-hunting results from frontier models Voices: Anthropic, Mozilla, OpenAI, Mandiant
Feb 5, 2026 Anthropic says Claude Opus 4.6 validated 500+ high-severity vulnerabilities security leader AI-based vulnerability discovery can be built into defensive workflows, but the vendor's claim needs independent verification.developer AI support can be effective in code review, but validating the findings still requires human work.journalist and researcher The 500+ figure is a vendor claim; the methodology awaits independent confirmation.Feb 5, 2026 AI discovers 12 vulnerabilities in OpenSSL security operations For systems that use OpenSSL, track the patches that are released.developer AI-based code analysis can produce results even in critical open-source libraries.journalist and researcher The results of AI-driven vulnerability discovery can be checked against the bug reports.Mar 6, 2026 Anthropic's Claude finds 22 security vulnerabilities in Firefox security leader On the defensive side, using AI to uncover flaws can be built into vulnerability management.developer AI-based bug hunting offers a new way to audit large codebases.Apr 7, 2026 Anthropic says Claude Mythos Preview found thousands of zero-days; Project Glasswing launched security leader Prepare for a wave of patches and faster discovery of vulnerabilities; the capabilities have not yet been independently verified.security operations Expect more patches for major systems, so patching processes should be sped up.developer Because AI can quickly find decades-old hidden bugs, code auditing is becoming more important.public-sector decision-maker The limited-access release of these capabilities raises questions about regulating AI capabilities and about defensive access to them.journalist and researcher Claims about the capabilities rest on Anthropic's own statements and need independent verification.Apr 21, 2026 Mozilla says Anthropic's Mythos found 271 security vulnerabilities in Firefox 150 security leader Using AI defensively in vulnerability discovery can speed up the patching cycle.developer AI-based vulnerability discovery can produce a large volume of findings, and handling them burdens the development process.Jun 24, 2026 US official says Anthropic's Mythos model found vulnerabilities in classified government systems security leader AI-based vulnerability discovery can be a real tool on the defensive side too, including in classified environments.public-sector decision-maker Flaws found in government systems raise the question of how AI tools should be introduced under regulation.Aug 17, 2026 Wiz AI agent finds Snowflake flaw that GitHub Copilot missed security leader Agent-based offensive testing can complement traditional assessment.developer Code-assistant review is no substitute for targeted security testing.Aug 18, 2026 Mandiant's agentic vulnerability discovery system finds over 100 critical flaws in two days security leader AI-based vulnerability discovery can speed up detection, but remediation capacity must grow too.developer Automated discovery may bring a larger volume of bug reports.Sep 28, 2026 Open-source AI security agent finds 24 Android vulnerabilities developer AI-based vulnerability scanning can provide a practical tool for code review.journalist and researcher The method and results are measurable examples of defensive use of agents.Nov 5, 2026 OpenAI AI pipeline generates 948 Linux root privilege-escalation exploits in 30 hours security leader Vulnerability management must account for the time available before an exploit appears, which AI may shorten further.security operations Patching Linux systems quickly and detecting privilege-escalation attempts matter even more.developer Automated exploit discovery raises the importance of security review in kernel and system-level code.journalist and researcher Without independent verification of the figure and the methodology, the real impact is hard to judge.Nov 18, 2026 AI used to uncover critical Java vulnerability, Euronews reports developer For Java-based products, watch for the fix to this flaw.journalist and researcher Another example of AI-based vulnerability research.Early access for vetted defenders Voices: OpenAI, Anthropic, White House, ENISA
Apr 14, 2026 OpenAI expands Trusted Access for Cyber and offers GPT-5.4-Cyber to vetted defenders security leader As AI capabilities grow, defenders may also need access programs.security operations The model, built for defense, is available through vetted access.Apr 16, 2026 White House to give US agencies access to Anthropic's Mythos security leader Government access shows that state defence, too, is counting on the model's capabilities.public-sector decision-maker US federal bodies are getting direct access to the model.Jun 1, 2026 Anthropic offers EU access to its Mythos hacking technology security leader Wider availability of AI-driven vulnerability discovery may change both defensive and threat models.public-sector decision-maker EU institutions must decide how to obtain access to an AI-based offensive capability and on what terms.Jun 17, 2026 EU cybersecurity agency to meet Anthropic on Thursday, Commission says security leader The meeting could set the direction for risk-management expectations towards AI providers.public-sector decision-maker Talks between EU regulators and agencies bear on how the security of AI models is handled.Jun 22, 2026 OpenAI's Patch the Planet initiative supports open source maintainers security leader AI-based vulnerability research could speed up the patching of open source dependencies.developer Open source maintainers may receive AI support in finding and fixing vulnerabilities.Jul 14, 2026 White House announces 'Gold Eagle' AI clearinghouse for cyber vulnerabilities security leader The discovery and handling of vulnerabilities may speed up.developer Developers of advanced AI may face expectations of cooperation.public-sector decision-maker The US model shows how early government access to AI capabilities can be organised.Aug 10, 2026 OpenAI expands Daybreak defence programme and launches new cybersecurity model security leader AI-based defensive tools could add a new layer to SOC work.security operations The new model's access terms and effectiveness are worth evaluating.Sep 10, 2026 EU cybersecurity agency ENISA granted access to Anthropic's Mythos 5 model, Commission says public-sector decision-maker Direct model access for the EU agency strengthens oversight of AI risks.journalist and researcher The terms and scope of the access are not clear from the announcement.Sep 22, 2026 OpenAI's Astra rollout limits cyber tools to 16 firms security leader Not everyone has access to the most capable AI-based defensive tools.journalist and researcher The access restriction raises questions about transparency.Sep 23, 2026 OpenAI extends Daybreak cyber defence tools to Ukraine for civilian protection security leader Defensive AI tools are becoming available in the affected region.public-sector decision-maker AI capabilities are also entering government cyber defence cooperation.Oct 6, 2026 Anthropic launches Cyber Verification Program and Cyber Mission initiative security operations Defence teams may gain verified access to the cybersecurity capabilities of AI models.journalist and researcher For researchers, the conditions of the verification program affect how far the models are available for research purposes.Overload of the disclosure and patching process Voices: NIST, ENISA, Google Project Zero, Anthropic
Apr 16, 2026 NIST moves National Vulnerability Database to risk-based CVE triage amid record submissions security leader Tools and processes should be reviewed for their dependence on the NVD.security operations Vulnerability management that relies on the NVD needs another source for CVEs that have not been enriched.journalist and researcher "Not Scheduled" CVEs do not disappear, but NIST will not enrich them automatically.Apr 20, 2026 NIST moves to risk-based NVD model as CVE submissions rise 263% since 2020 security leader Organisations will increasingly have to prioritise vulnerabilities and score them themselves.security operations Vulnerability management tools can no longer count on NVD enrichment and scores for every CVE.Aug 5, 2026 AI slop pollutes the CVE process with fake vulnerabilities security operations Fake CVE entries add to the burden of prioritisation, so sources need to be verified.developer Maintainers may be swamped by a flood of fake AI-generated reports.journalist and researcher Screening and authenticating vulnerability reports is becoming more important.Aug 10, 2026 ENISA expands CVE Root to 20 CNAs security operations Coordinated vulnerability disclosure can rely on a wider European network.developer The expansion of the CNA network affects how vulnerabilities are reported.Oct 21, 2026 magyarorszag.hu and government sites in Slovakia go offline over PHP flaw security operations Deliberately shutting down the government portal and restarting it safely after patching is their task.public-sector decision-maker The outage affects electronic administration, so decisions are also needed on communication and interim arrangements.citizen While the portal is down, online administration is not available there.Oct 21, 2026 AI-discovered critical PHP flaw takes government websites offline worldwide security leader Must immediately assess where PHP runs across the estate and speed up patching.security operations Responsible for urgent patching and monitoring logs for attack attempts.developer Applications built on PHP must be updated and tested against the patched version.business leader Websites and services running PHP carry a risk of outages and attacks.public-sector decision-maker Some government sites were taken offline because of the flaw, so decisions are also needed on the outage and on recovery.Nov 15, 2026 Google Project Zero to shorten vulnerability disclosure deadline to 30 days by 2028 security operations Patches and public details will follow each other more quickly.developer Vendors have less time to produce a fix before public disclosure.journalist and researcher The change in public disclosure practice also affects researchers' disclosure timelines.Nov 16, 2026 Anthropic and Google Project Zero both shorten vulnerability disclosure deadlines security operations Details become public sooner, so a faster patch cycle is needed.developer Vendors must release fixes in less time, or the flaw may be disclosed publicly.journalist and researcher Changes in disclosure practice and in the frameworks of coordinated responsible disclosure need to be tracked.Who will get it first: an attacker race Voices: Google, Amazon, Anthropic, VulnCheck
Jan 23, 2026 Heise reports that AI-generated zero-day exploits are already a reality security leader AI can speed up the exploitation of vulnerabilities, so patching windows should be shortened.journalist and researcher The technical basis for the claim should be checked in the source article.May 11, 2026 Google detects AI-developed zero-day exploit tied to cybercriminal group security leader AI-accelerated vulnerability research and exploit development mean shorter defence windows, which should be factored into threat models and patch priorities.security operations Because of the reported bypass of two-factor protection, it is worth reviewing how authentication processes are monitored.developer Because AI can uncover flaws, code auditing and rapid patching matter even more.journalist and researcher The case is a defender-documented example of AI-assisted exploit development that researchers should follow.Jul 29, 2026 VulnCheck: only 1% of AI-discovered vulnerabilities exploited in the wild security leader AI-driven vulnerability research may increase the patching workload, while the exploitation rate remains low for now.journalist and researcher The measured rate adds data to the debate on AI's offensive capabilities.Sep 10, 2026 Anthropic: AI lets smaller actors run state-level hacking campaigns security leader The report says AI appears in attacks at varying levels of autonomy, which affects threat modelling.security operations Anthropic says attacker sophistication no longer depends on the size of the actor, so detection assumptions need rethinking.journalist and researcher The report gives details of cases ranging from cyberattacks to influence operations, surveillance, biology and weapons building.Nov 16, 2026 AI can now find zero-day flaws for as little as $4, report says security leader As the cost of finding vulnerabilities falls, shortening the time to patch becomes more important.developer Defenders can also use AI tools to scan code for security flaws.Nov 16, 2026 Google: attackers use AI agents to turn software patches into working exploits security leader Patch-management SLAs should be re-evaluated in light of automated exploit development.security operations The window between a patch release and an exploit appearing is narrowing, so patching speed matters even more.developer Published security fixes could be quickly reverse-engineered and exploited by AI agents.Nov 20, 2026 Amazon says Chinese-nexus hackers used Cairn AI tool to find apparent zero-day in attack on university security leader AI-assisted zero-day discovery narrows defenders' advantage over unknown flaws.journalist and researcher University and research environments can be prime targets for AI-driven espionage.Who is affected and how security operations Following the PHP case pattern, an AI-discovered flaw can cause mass outages overnight. Maintain an up-to-date inventory of runtime environments to identify affected systems within hours. developer Prepare for a shorter disclosure window: the 30-day deadline for severe flaws also pressures the patching cycle of your own product. It is advisable to run AI-based code checks before release. public-sector decision-maker Government service outages are a realistic scenario: in the magyarorszag.hu case, up to a 48-hour processing suspension had to be anticipated. A pre-agreed communication and alternative processing plan is required. journalist and researcher Results from closed programs cannot be independently verified, so seek Mozilla-type figures validated by third parties. Hungary The only domestic trace is the shutdown of magyarorszag.hu due to a PHP bug affecting multiple countries, potentially causing a 48-hour service interruption. This indicates that Hungarian state systems are directly impacted by the international bug discovery wave.
Weekly trend Supporting stories per week
2 Nov 17–23: 0 Nov 24–30: 0 dec. 1–7.: 0 dec. 8–14.: 0 dec. 15–21.: 0 dec. 22–28.: 0 Dec 29 – Jan 4: 0 jan. 5–11.: 0 jan. 12–18.: 0 jan. 19–25.: 0 Jan 26 – Feb 1: 0 febr. 2–8.: 0 febr. 9–15.: 0 febr. 16–22.: 0 Feb 23 – Mar 1: 0 Mar 2–8: 1 Mar 9–15: 0 Mar 16–22: 0 Mar 23–29: 0 Mar 30 – Apr 5: 0 Apr 6–12: 1 Apr 13–19: 0 Apr 20–26: 0 Apr 27 – May 3: 0 May 4–10: 0 May 11–17: 0 May 18–24: 0 May 25–31: 0 Jun 1–7: 0 Jun 8–14: 0 Jun 15–21: 0 Jun 22–28: 1 Jun 29 – Jul 5: 0 Jul 6–12: 0 Jul 13–19: 0 Jul 20–26: 0 Jul 27 – Aug 2: 0 Aug 3–9: 0 Aug 10–16: 0 Aug 17–23: 0 Aug 24–30: 0 Aug 31 – Sep 6: 0 Sep 7–13: 1 Sep 14–20: 0 Sep 21–27: 1 Sep 28 – Oct 4: 0 Oct 5–11: 0 Oct 12–18: 0 Oct 19–25: 2 Oct 26 – Nov 1: 0 Nov 2–8: 1 Nov 9–15: 0 Nov 16–22: 2 Nov 17–23 Nov 16–22
Turning points Related narratives in the News magazine Supporting stories Mar 6, 2026 Anthropic's Claude finds 22 security vulnerabilities in Firefox security leader On the defensive side, using AI to uncover flaws can be built into vulnerability management.developer AI-based bug hunting offers a new way to audit large codebases.Apr 7, 2026 Anthropic says Claude Mythos Preview found thousands of zero-days; Project Glasswing launched security leader Prepare for a wave of patches and faster discovery of vulnerabilities; the capabilities have not yet been independently verified.security operations Expect more patches for major systems, so patching processes should be sped up.developer Because AI can quickly find decades-old hidden bugs, code auditing is becoming more important.public-sector decision-maker The limited-access release of these capabilities raises questions about regulating AI capabilities and about defensive access to them.journalist and researcher Claims about the capabilities rest on Anthropic's own statements and need independent verification.Jun 24, 2026 US official says Anthropic's Mythos model found vulnerabilities in classified government systems security leader AI-based vulnerability discovery can be a real tool on the defensive side too, including in classified environments.public-sector decision-maker Flaws found in government systems raise the question of how AI tools should be introduced under regulation.Sep 10, 2026 Anthropic: AI lets smaller actors run state-level hacking campaigns security leader The report says AI appears in attacks at varying levels of autonomy, which affects threat modelling.security operations Anthropic says attacker sophistication no longer depends on the size of the actor, so detection assumptions need rethinking.journalist and researcher The report gives details of cases ranging from cyberattacks to influence operations, surveillance, biology and weapons building.Sep 22, 2026 AI-guided malware: self-rewriting code, CLOSEDQUORUM and Cisco Talos's CAIRN tool security operations New detection approaches are needed for shifting malware that uses AI.journalist and researcher CAIRN is open source, so researchers can try it and extend it.Oct 21, 2026 AI-discovered critical PHP flaw takes government websites offline worldwide security leader Must immediately assess where PHP runs across the estate and speed up patching.security operations Responsible for urgent patching and monitoring logs for attack attempts.developer Applications built on PHP must be updated and tested against the patched version.business leader Websites and services running PHP carry a risk of outages and attacks.public-sector decision-maker Some government sites were taken offline because of the flaw, so decisions are also needed on the outage and on recovery.Oct 21, 2026 magyarorszag.hu and government sites in Slovakia go offline over PHP flaw security operations Deliberately shutting down the government portal and restarting it safely after patching is their task.public-sector decision-maker The outage affects electronic administration, so decisions are also needed on communication and interim arrangements.citizen While the portal is down, online administration is not available there.Nov 5, 2026 OpenAI AI pipeline generates 948 Linux root privilege-escalation exploits in 30 hours security leader Vulnerability management must account for the time available before an exploit appears, which AI may shorten further.security operations Patching Linux systems quickly and detecting privilege-escalation attempts matter even more.developer Automated exploit discovery raises the importance of security review in kernel and system-level code.journalist and researcher Without independent verification of the figure and the methodology, the real impact is hard to judge.Nov 16, 2026 Anthropic and Google Project Zero both shorten vulnerability disclosure deadlines security operations Details become public sooner, so a faster patch cycle is needed.developer Vendors must release fixes in less time, or the flaw may be disclosed publicly.journalist and researcher Changes in disclosure practice and in the frameworks of coordinated responsible disclosure need to be tracked.Nov 16, 2026 Google: attackers use AI agents to turn software patches into working exploits security leader Patch-management SLAs should be re-evaluated in light of automated exploit development.security operations The window between a patch release and an exploit appearing is narrowing, so patching speed matters even more.developer Published security fixes could be quickly reverse-engineered and exploited by AI agents.Machine analysis based on story titles and summaries, with a verification pass. The source of the facts is the article linked at each story.
What do you make of this piece?
I agree I'm skeptical The split shows after 20 votes.
Why? Your answer is not shown publicly; it helps the editors.
Well supported Matches my experience Brought a new angle Important to cover Thin or weak sourcing One-sided Doesn't match my experience Overstated Outdated
Saved pieces Weekly email One email a week with the main stories from the Agent Age editions. You confirm by email and can unsubscribe in one click.
Almost done: we sent you an email. Open it and press the confirm link. If it does not arrive within a few minutes, check your spam folder.
Please enter a valid email address and tick the consent box.
Something went wrong. Please try again later.