2 risks moving Reports a month, each counted once: ○ Feb–Jul ● Aug–Oct
Past flags that held the next month: 2 of 5 (40%)
Incidents by type of risk · Last 3 months
How we count 273 real-world incidents made public in Aug 2026 – Oct 2026, by month, filed under their main risk. Lab tests apart (45): incidents whose source states an evaluation, training or research setting. An incident counts once.
Aug Sep Oct Total
All incidents
43 67 163
273 AI used to run a cyber attack
7 13 31
51 Hidden instructions hijack the agent
6 9 15
30 Harmful content and jailbreaks
9 2 15
26 Agent takes destructive or unapproved actions
6 2 13
21 AI system attacked or breached
2 3 13
18 Fraud and impersonation
4 5 8
17 Credential and privilege abuse
6 10
16 Manipulation and misinformation
3 6 7
16 Physical harm
2 4 10
16 Made-up output relied on
3 4 7
14 Privacy harm to a person
1 2 8
11 Data leaked or sent out
5 5
10 Psychological harm and self-harm
1 5
6 Agent escapes its sandbox
1 3
4 Bias and discrimination
3
3 Copyright and IP
1 2
3 Supply-chain compromise
1 2
3 No risk filed
1 2
3 Deception, concealment or sabotage
2
2 Agents coordinating against their limits
1
1 Consumer detriment and unsuitable advice
1
1 Silent wrong state or drift
1
1
Each risk's share of the month's incidents % of that month's incidents
Dec Feb Mar Apr May Jun Jul Aug Sep Oct
AI used to run a cyber attack 1 of 31 incidents 3 4 of 50 incidents 8 3 of 47 incidents 6 2 of 48 incidents 4 1 of 34 incidents 3 3 of 45 incidents 7 11 of 66 incidents 17 7 of 57 incidents 12 13 of 77 incidents 17 32 of 184 incidents 17 Harmful content and jailbreaks 3 of 31 incidents 10 5 of 50 incidents 10 4 of 47 incidents 9 7 of 48 incidents 15 2 of 34 incidents 6 4 of 45 incidents 9 9 of 66 incidents 14 9 of 57 incidents 16 2 of 77 incidents 3 15 of 184 incidents 8 Agent takes destructive or unapproved actions 2 of 31 incidents 6 6 of 50 incidents 12 7 of 47 incidents 15 5 of 48 incidents 10 3 of 34 incidents 9 3 of 45 incidents 7 4 of 66 incidents 6 7 of 57 incidents 12 5 of 77 incidents 6 16 of 184 incidents 9 Hidden instructions hijack the agent 1 of 31 incidents 3 3 of 50 incidents 6 1 of 47 incidents 2 2 of 48 incidents 4 2 of 34 incidents 6 9 of 45 incidents 20 7 of 66 incidents 11 7 of 57 incidents 12 9 of 77 incidents 12 15 of 184 incidents 8 Physical harm 4 of 31 incidents 13 3 of 50 incidents 6 6 of 47 incidents 13 6 of 48 incidents 12 3 of 34 incidents 9 3 of 45 incidents 7 5 of 66 incidents 8 2 of 57 incidents 4 4 of 77 incidents 5 10 of 184 incidents 5 Made-up output relied on 3 of 31 incidents 10 4 of 50 incidents 8 6 of 47 incidents 13 5 of 48 incidents 10 4 of 34 incidents 12 4 of 45 incidents 9 2 of 66 incidents 3 3 of 57 incidents 5 4 of 77 incidents 5 8 of 184 incidents 4 AI system attacked or breached 2 of 31 incidents 6 6 of 50 incidents 12 4 of 47 incidents 9 4 of 48 incidents 8 2 of 34 incidents 6 2 of 45 incidents 4 2 of 66 incidents 3 2 of 57 incidents 4 3 of 77 incidents 4 13 of 184 incidents 7 Manipulation and misinformation 2 of 31 incidents 6 0 of 50 incidents 0 2 of 47 incidents 4 5 of 48 incidents 10 3 of 34 incidents 9 5 of 45 incidents 11 3 of 66 incidents 5 3 of 57 incidents 5 6 of 77 incidents 8 8 of 184 incidents 4 Privacy harm to a person 5 of 31 incidents 16 7 of 50 incidents 14 1 of 47 incidents 2 2 of 48 incidents 4 1 of 34 incidents 3 2 of 45 incidents 4 4 of 66 incidents 6 1 of 57 incidents 2 2 of 77 incidents 3 8 of 184 incidents 4 Fraud and impersonation 3 of 31 incidents 10 5 of 50 incidents 10 2 of 47 incidents 4 0 of 48 incidents 0 2 of 34 incidents 6 0 of 45 incidents 0 2 of 66 incidents 3 4 of 57 incidents 7 5 of 77 incidents 6 8 of 184 incidents 4 Credential and privilege abuse 0 of 31 incidents 0 2 of 50 incidents 4 1 of 47 incidents 2 3 of 48 incidents 6 2 of 34 incidents 6 0 of 45 incidents 0 0 of 66 incidents 0 1 of 57 incidents 2 8 of 77 incidents 10 13 of 184 incidents 7 Data leaked or sent out 0 of 31 incidents 0 0 of 50 incidents 0 1 of 47 incidents 2 1 of 48 incidents 2 1 of 34 incidents 3 3 of 45 incidents 7 0 of 66 incidents 0 0 of 57 incidents 0 7 of 77 incidents 9 6 of 184 incidents 3
Which risks are growing as a share of incidents · Last 3 months
How we count Each risk's share of real-world incidents, May 2026 – Jul 2026 (open circle) against Aug 2026 – Oct 2026 (filled), from the publishers already recorded in the earlier period: 137 and 139 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.
○ May 2026 – Jul 2026 ● Aug 2026 – Oct 2026
0% 5% 10% 15%
Credential and privilege abuse
1.5%
7.2% AI used to run a cyber attack
11%
18% AI system attacked or breached
4.4%
7.2% Physical harm
8.0%
9.4% Hidden instructions hijack the agent
12%
13% Agent takes destructive or unapproved actions
6.6%
7.2% Harmful content and jailbreaks
7.9%
11% Agents coordinating against their limits
0%
0.7% Deception, concealment or sabotage
0%
0.7% Copyright and IP
0.7%
1.4% Data leaked or sent out
2.2%
2.9% Fraud and impersonation
2.9% Agent escapes its sandbox
0%
0.7% Unauthorised transactions or spend
0%
0.7% Psychological harm and self-harm
2.2%
2.9% Silent wrong state or drift
0%
1.5% Consumer detriment and unsuitable advice
0.7%
2.2% Bias and discrimination
2.2%
3.6% Privacy harm to a person
3.6%
5.1% Provider concentration or outage
0%
2.2% Supply-chain compromise
2.2%
4.4% Manipulation and misinformation
5.8%
8.0% Made-up output relied on
4.3%
7.3%
Starts with Breaking in, 15 Products ARTEX 4 · Hermes Agent 3 Most often after Affected people told, 10
Starts with Breaking in, 12 Most often after Access removed, 8
Risks seen for the first time · Last 3 months
How we count Risks whose first real-world incident on record was made public in Aug 2026 – Oct 2026, with the date of that first incident and the incidents in the period.
Agents coordinating against their limits (first seen 4 Sep 2026)
1 Deception, concealment or sabotage (first seen 23 Oct 2026)
2
What labs test for, against what happens in real use · Last 3 months
How we count For each risk: checked safety test results for the model versions released in Aug 2026 – 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 Aug 2026 – Oct 2026. Each column has its own scale.
Safety test results Real-world incidents
Deception, concealment or sabotage
295 2 Harmful content and jailbreaks
226 26 AI used to run a cyber attack
187 51 Weapons uplift (CBRN)
176 Hidden instructions hijack the agent
66 30 Agent escapes its sandbox
46 4 AI R&D acceleration
15 Gaming the task (reward hacking)
13 Agents coordinating against their limits
10 1 Manipulation and misinformation
6 16 Agent takes destructive or unapproved actions
5 21 Credential and privilege abuse
4 16 Data leaked or sent out
3 10 AI system attacked or breached
18 Fraud and impersonation
17 Physical harm
16 Made-up output relied on
14 Privacy harm to a person
11 Psychological harm and self-harm
6 Bias and discrimination
3 Copyright and IP
3 Supply-chain compromise
3 Consumer detriment and unsuitable advice
1 Silent wrong state or drift
1