AI security
AI threat tracker
How attacks on AI software, new vulnerabilities in AI packages, reported AI incidents and threat actors’ use of AI change month by month, from public sources.
Updated 21 Nov 2026
AI software under active attack
CISA added 10 vulnerabilities in AI software to its list of exploited vulnerabilities in the last 12 months, up from 1 in the 12 months before.
Show the numbers
| Month | added |
|---|---|
| Dec 2024 | 0 |
| Jan 2025 | 0 |
| Feb 2025 | 0 |
| Mar 2025 | 0 |
| Apr 2025 | 0 |
| May 2025 | 1 |
| Jun 2025 | 0 |
| Jul 2025 | 0 |
| Aug 2025 | 0 |
| Sept 2025 | 0 |
| Oct 2025 | 0 |
| Nov 2025 | 0 |
| Dec 2025 | 0 |
| Jan 2026 | 0 |
| Feb 2026 | 0 |
| Mar 2026 | 1 |
| Apr 2026 | 0 |
| May 2026 | 2 |
| Jun 2026 | 1 |
| Jul 2026 | 2 |
| Aug 2026 | 2 |
| Sept 2026 | 1 |
| Oct 2026 | 0 |
| Nov 2026 | 1 |
By kind of software
| Kind of software | Last 12 months | 12 months before |
|---|---|---|
| Agent builders and frameworks | 5 | 1 |
| AI apps, gateways and interfaces | 4 | 0 |
| ML operations and data tools | 1 | 0 |
Latest additions
What this does not show
The list only includes vulnerabilities CISA has reliable evidence are being exploited, and it leans towards software used by US government agencies.
Hosted AI services such as chatbots rarely get CVE numbers, so attacks on them do not appear here.
An entry is counted when the product is on ResearchNode’s list of AI software. Entries that only mention AI in their description are checked by hand first.
New vulnerabilities in AI packages
551 vulnerabilities were published for the 62 AI packages on the watchlist in the last 12 months, up from 222 the year before. 332 are rated critical or high, and 14 have no rating yet.
new vulnerabilitiescritical or high
Show the numbers
| Month | new vulnerabilities | critical or high |
|---|---|---|
| Dec 2024 | 1 | 1 |
| Jan 2025 | 3 | 2 |
| Feb 2025 | 3 | 1 |
| Mar 2025 | 83 | 59 |
| Apr 2025 | 16 | 8 |
| May 2025 | 16 | 8 |
| Jun 2025 | 5 | 4 |
| Jul 2025 | 21 | 11 |
| Aug 2025 | 8 | 6 |
| Sept 2025 | 30 | 21 |
| Oct 2025 | 22 | 13 |
| Nov 2025 | 12 | 10 |
| Dec 2025 | 25 | 24 |
| Jan 2026 | 15 | 15 |
| Feb 2026 | 23 | 15 |
| Mar 2026 | 45 | 36 |
| Apr 2026 | 58 | 38 |
| May 2026 | 110 | 73 |
| Jun 2026 | 62 | 29 |
| Jul 2026 | 44 | 22 |
| Aug 2026 | 53 | 35 |
| Sept 2026 | 49 | 14 |
| Oct 2026 | 64 | 28 |
| Nov 2026 | 1 | 1 |
By kind of software
| Kind of software | Last 12 months | 12 months before | Most often affected |
|---|---|---|---|
| AI apps, gateways and interfacesopen-webui · 139 | 177 | 39 | open-webui · 139 |
| Agent builders and frameworksflowise · 81 | 140 | 49 | flowise · 81 |
| Model servers and inferencevllm · 82 | 118 | 50 | vllm · 82 |
| ML frameworks and librariestransformers · 13 | 43 | 47 | transformers · 13 |
| ML operations and data toolsmlflow · 32 | 39 | 26 | mlflow · 32 |
| MCP servers and SDKsmcp · 9 | 25 | 10 | mcp · 9 |
| Vector databaseschromadb · 4 | 9 | 1 | chromadb · 4 |
Latest critical vulnerabilities
What this does not show
More vulnerabilities can mean more people are looking, not that the software got worse. The exploited list above is the better sign of real attacks.
Only packages published on PyPI, npm and Go are counted. AI apps installed from source, and hosted services, are missing.
The same flaw listed by GitHub, PyPI and NVD is counted once, by its CVE number. Severity is GitHub’s reviewed rating where there is one.
Reported AI incidents
The AI Incident Database lists 373 incidents that happened in the last 12 months, down from 446 in the 12 months before. 72 of them are classified as malicious actors and misuse, against 234 the year before.
Source: Responsible AI Collaborative’s AI Incident Database export from 16 Nov 2026. ResearchNode groups the records by incident date and MIT risk category. Incident data: CC BY-SA 4.0.
The dashed end and open nodes are the last three months, which are still being filled in. Incidents are often added weeks after they happen.
incidentsmalicious actors and misuse
Show the numbers
| Month | incidents | malicious actors and misuse |
|---|---|---|
| Dec 2024 | 26 | 18 |
| Jan 2025 | 33 | 19 |
| Feb 2025 | 47 | 30 |
| Mar 2025 | 46 | 26 |
| Apr 2025 | 48 | 27 |
| May 2025 | 34 | 18 |
| Jun 2025 | 45 | 25 |
| Jul 2025 | 39 | 15 |
| Aug 2025 | 35 | 16 |
| Sept 2025 | 22 | 9 |
| Oct 2025 | 42 | 19 |
| Nov 2025 | 33 | 14 |
| Dec 2025 | 45 | 19 |
| Jan 2026 | 39 | 17 |
| Feb 2026 | 33 | 9 |
| Mar 2026 | 29 | 11 |
| Apr 2026 | 29 | 7 |
| May 2026 | 42 | 5 |
| Jun 2026 | 32 | 0 |
| Jul 2026 | 39 | 0 |
| Aug 2026 | 17 | 0 |
| Sept 2026 | 30 | 0 |
| Oct 2026 | 25 | 0 |
| Nov 2026 | 4 | 0 |
By risk area, last 24 months
| Risk area (MIT AI Risk Repository) | Incidents | Share |
|---|---|---|
| Malicious actors and misuse | 306 | 37% |
| Not yet classified | 279 | 34% |
| Misinformation | 85 | 10% |
| AI system safety, failures and limitations | 54 | 7% |
| Privacy and security | 36 | 4% |
| Discrimination and toxicity | 34 | 4% |
| Human-computer interaction | 19 | 2% |
| Socioeconomic and environmental | 6 | 1% |
Most recent incidents
What this does not show
Reported is not the same as happened. The count follows what news media and submitters notice, and media attention to AI keeps growing.
The database is maintained by volunteer editors, so its coverage depends on what they review.
Risk areas come from the database’s own MIT classification, which is partly done by a language model.
How threat actors use AI
No records yet. Records appear here once a report has been read and checked.
What this does not show
Each company only sees misuse of its own AI products, and publishes on its own schedule. A jump usually means a report came out, so this is shown as records, not a trend line.
Names, countries and attributions are the publishing company’s.
Sources
The tracker updates automatically from these sources. Spotted something wrong? See Corrections.