AI Risk Wire

AI Risk Monitor

2 risks moving

Reports a month, each counted once: ○ Feb–Jul ● Aug–Oct

RiskIncidents per month, last 12 monthsReports a month, before → nowLab reports nowReading
AI used to run a cyber attack4.0 → 9.71Rising
Hidden instructions hijack the agent3.7 → 6.31Not a trend
Physical harm4.3 → 4.7Not a trend
Harmful content and jailbreaks5.2 → 4.0Not a trend
Agent takes destructive or unapproved actions4.5 → 3.77Not a trend
Credential and privilege abuse1.2 → 3.76Rising
AI system attacked or breached3.3 → 3.7Not a trend
Manipulation and misinformation3.0 → 3.01Not a trend

Past flags that held the next month: 2 of 5 (40%)

Incidents by type of risk · Last month

How we count

163 real-world incidents made public in Oct 2026, by week, filed under their main risk. Lab tests apart (21): incidents whose source states an evaluation, training or research setting. An incident counts once.

1–4 Oct5–11 Oct12–18 Oct19–25 Oct26–31 OctTotal All incidents 1724333158 163AI used to run a cyber attack 221818 31Harmful content and jailbreaks 22335 15Hidden instructions hijack the agent 24324 15AI system attacked or breached 22315 13Agent takes destructive or unapproved actions 3325 13Credential and privilege abuse 1243 10Physical harm 1612 10Fraud and impersonation 2213 8Privacy harm to a person 11312 8Made-up output relied on 2131 7Manipulation and misinformation 2131 7Data leaked or sent out 11111 5Psychological harm and self-harm 1121 5Agent escapes its sandbox 111 3Bias and discrimination 12 3Copyright and IP 2 2Deception, concealment or sabotage 11 2Supply-chain compromise 2 2No risk filed 11 2Consumer detriment and unsuitable advice 1 1Silent wrong state or drift 1 1

Each risk's share of the month's incidents

% of that month's incidents

DecFebMarAprMayJunJulAugSepOct AI used to run a cyber attack1 of 31 incidents34 of 50 incidents83 of 47 incidents62 of 48 incidents41 of 34 incidents33 of 45 incidents711 of 66 incidents177 of 57 incidents1213 of 77 incidents1732 of 184 incidents17Harmful content and jailbreaks3 of 31 incidents105 of 50 incidents104 of 47 incidents97 of 48 incidents152 of 34 incidents64 of 45 incidents99 of 66 incidents149 of 57 incidents162 of 77 incidents315 of 184 incidents8Agent takes destructive or unapproved actions2 of 31 incidents66 of 50 incidents127 of 47 incidents155 of 48 incidents103 of 34 incidents93 of 45 incidents74 of 66 incidents67 of 57 incidents125 of 77 incidents616 of 184 incidents9Hidden instructions hijack the agent1 of 31 incidents33 of 50 incidents61 of 47 incidents22 of 48 incidents42 of 34 incidents69 of 45 incidents207 of 66 incidents117 of 57 incidents129 of 77 incidents1215 of 184 incidents8Physical harm4 of 31 incidents133 of 50 incidents66 of 47 incidents136 of 48 incidents123 of 34 incidents93 of 45 incidents75 of 66 incidents82 of 57 incidents44 of 77 incidents510 of 184 incidents5Made-up output relied on3 of 31 incidents104 of 50 incidents86 of 47 incidents135 of 48 incidents104 of 34 incidents124 of 45 incidents92 of 66 incidents33 of 57 incidents54 of 77 incidents58 of 184 incidents4AI system attacked or breached2 of 31 incidents66 of 50 incidents124 of 47 incidents94 of 48 incidents82 of 34 incidents62 of 45 incidents42 of 66 incidents32 of 57 incidents43 of 77 incidents413 of 184 incidents7Manipulation and misinformation2 of 31 incidents60 of 50 incidents02 of 47 incidents45 of 48 incidents103 of 34 incidents95 of 45 incidents113 of 66 incidents53 of 57 incidents56 of 77 incidents88 of 184 incidents4Privacy harm to a person5 of 31 incidents167 of 50 incidents141 of 47 incidents22 of 48 incidents41 of 34 incidents32 of 45 incidents44 of 66 incidents61 of 57 incidents22 of 77 incidents38 of 184 incidents4Fraud and impersonation3 of 31 incidents105 of 50 incidents102 of 47 incidents40 of 48 incidents02 of 34 incidents60 of 45 incidents02 of 66 incidents34 of 57 incidents75 of 77 incidents68 of 184 incidents4Credential and privilege abuse0 of 31 incidents02 of 50 incidents41 of 47 incidents23 of 48 incidents62 of 34 incidents60 of 45 incidents00 of 66 incidents01 of 57 incidents28 of 77 incidents1013 of 184 incidents7Data leaked or sent out0 of 31 incidents00 of 50 incidents01 of 47 incidents21 of 48 incidents21 of 34 incidents33 of 45 incidents70 of 66 incidents00 of 57 incidents07 of 77 incidents96 of 184 incidents3

Which risks are growing as a share of incidents · Last month

How we count

Each risk's share of real-world incidents, Sep 2026 (open circle) against Oct 2026 (filled), from the publishers already recorded in the earlier period: 67 and 80 incidents. Dark: rising, passes the trend test. Mid grey: falling, passes it. Light grey: tested, no change that passes. Pale: fewer than 10 incidents in the later period, lab tests included, so not tested. Where our records have months with no real-world incident before the period, the longest stretch they allow is compared.

○ Sep 2026● Oct 2026
0%5%10%15%20% AI used to run a cyber attack 19% 20%Credential and privilege abuse 8.8% 9.0%Harmful content and jailbreaks 3.0% 8.8%AI system attacked or breached 4.5% 10%Bias and discrimination 0% 3.8%Agent takes destructive or unapproved actions 3.0% 6.3%Deception, concealment or sabotage 0% 2.5%Consumer detriment and unsuitable advice 0% 1.3%Copyright and IP 1.5% 2.5%Psychological harm and self-harm 1.5% 2.5%Supply-chain compromise 1.5% 2.5%Privacy harm to a person 3.0% 3.8%Physical harm 6.0% 6.3%Agent escapes its sandbox 1.3% 1.5%Agents coordinating against their limits 0% 1.5%Made-up output relied on 3.8% 6.0%Hidden instructions hijack the agent 10% 13%Data leaked or sent out 1.3% 7.5%Fraud and impersonation 1.3% 7.5%Manipulation and misinformation 2.5% 9.0%

AI used to run a cyber attack

Starts with
Breaking in, 15
Products
ARTEX 4 · Hermes Agent 3
Most often after
Affected people told, 10

Credential and privilege abuse

Starts with
Breaking in, 12
Most often after
Access removed, 8

When each risk first appeared in our records · Last month

How we count

The first real-world incident on record for each risk that has incidents in Oct 2026, with its incidents in the period. Our records start 2 Nov 2024, so a risk first seen soon after may be older.

Psychological harm and self-harm (first seen 2 Nov 2024) 5Agent takes destructive or unapproved actions (first seen 21 Jul 2025) 13Supply-chain compromise (first seen 23 Jul 2025) 2AI used to run a cyber attack (first seen 27 Aug 2025) 31Manipulation and misinformation (first seen 2 Nov 2025) 7Hidden instructions hijack the agent (first seen 3 Nov 2025) 15Made-up output relied on (first seen 4 Nov 2025) 7Bias and discrimination (first seen 5 Nov 2025) 3Privacy harm to a person (first seen 7 Nov 2025) 8AI system attacked or breached (first seen 12 Nov 2025) 13Harmful content and jailbreaks (first seen 12 Nov 2025) 15Copyright and IP (first seen 12 Nov 2025) 2Physical harm (first seen 21 Nov 2025) 10Fraud and impersonation (first seen 1 Dec 2025) 8Credential and privilege abuse (first seen 2 Feb 2026) 10Silent wrong state or drift (first seen 8 Feb 2026) 1Data leaked or sent out (first seen 3 Mar 2026) 5Consumer detriment and unsuitable advice (first seen 12 May 2026) 1Agent escapes its sandbox (first seen 23 Jul 2026) 3Deception, concealment or sabotage (first seen 23 Oct 2026) 2

6 other risks had no incident in this period.

What labs test for, against what happens in real use · Last month

How we count

For each risk: checked safety test results for the model versions released in Oct 2026 (a result copied on two pages counts once; a result filed under two risks counts under both), and real-world incidents made public in Oct 2026. Each column has its own scale.

Safety test resultsReal-world incidents Deception, concealment or sabotage 1372Harmful content and jailbreaks 8915Weapons uplift (CBRN) 88AI used to run a cyber attack 7231Hidden instructions hijack the agent 3015Agent escapes its sandbox 213AI R&D acceleration 7Agent takes destructive or unapproved actions 413Gaming the task (reward hacking) 4Manipulation and misinformation 37Agents coordinating against their limits 3Data leaked or sent out 15AI system attacked or breached 13Credential and privilege abuse 10Physical harm 10Fraud and impersonation 8Privacy harm to a person 8Made-up output relied on 7Psychological harm and self-harm 5Bias and discrimination 3Copyright and IP 2Supply-chain compromise 2Consumer detriment and unsuitable advice 1Silent wrong state or drift 1

Built 16 Nov 2026. Periods end with the last full week or month; the week or month in progress is not counted. Bold: the five largest rows in each chart, ties included; controls by incidents where the control broke. Linked numbers open the incidents or model versions they count.

Reported incidents, not rates · 75 sources polled hourly since Sep 2026; earlier incidents come from a one-off backfill · English only · stories told twice are not yet merged · our own severity grades are not counted · built 2026-11-16T15:22:29Z

Linked numbers open the incidents they count. Updated 2026-11-16T15:22:29Z