11th AI Barometer – August 24 to 30, 2026

All AI Barometers

Observation Level: 🟢 Green

Across 4,800 executions distributed among 12 systems, AI Barometer #11 remains largely in the normal regime, with average stability virtually unchanged around 0.923.

Global semantic variation reaches 1.93%, compared with 1.57% in Barometer #10. Factual alerts remain limited, and no generalized shift of the panel is observed.

Reading: global stability maintained, with a slight rise in observed variations and a few localized movements to follow over time.

Weekly cartography

Observation Level: 🟢 Green

This week, the cartography of AI Barometer #11 remains at the green level.

The NORMAL_SIGNAL regime remains largely dominant, with 4,671 executions out of 4,800, or 97.31% of the panel.

Average stability remains virtually unchanged around 0.923, while global semantic variation rises slightly, from 1.57% in Barometer #10 to 1.93% this week.

The other regimes remain minority:

  • 91 observations in SEMANTIC_VARIATION
  • 14 in FACTUAL_ALERT
  • 1 in COMBINED_ALERT
  • 23 incomplete measurements

At profile level, NORMAL_SIGNAL remains the dominant regime for all 12 observed systems. Several localized shifts nevertheless appear, notably PROFILE-212079 at 5.26% semantic variation, while other profiles remain very close to their usual regime.

The cartography therefore does not show a general break in the panel. It rather describes a still globally stable set, within which a few local variations remain visible and will continue to be followed longitudinally.

NeoMundi weekly barometer #11

12
de-identified profiles
4,800
executions
99.521 %
coverage
92.29 %
mean stability
1.93 %
semantic variation

NeoMundi Barometer · week 11 · 12 de-identified profiles · generated 30/08/2026 18:47

Mapping of 12 de-identified observed systems. Horizontal axis: responses needing more verification. Vertical axis: change in response meaning. Graphic specification comparable across all weeks.

0 0.5 1 1.5 2 2.5 0 2 4 6 8 10 12 Responses needing more verification (%) the higher this number, the more responses need to be checked Change in response meaning (%) the meaning of responses changes from one run to another Meaning often changes, few responses need verification Meaning remains stable, more responses need verification Empty corner : no profile combines major meaning changes and more verification needs

Metrics table

A concise view of the main signals for each de-identified profile.

ProfileexecutionsFully scoredcoveragemean stabilitysemantic variationFLAG rateDominant regime
PROFILE-161CC540039999.750 %92.21 %0.25 %0.25 %NORMAL_SIGNAL
PROFILE-21207940039999.750 %92.31 %5.26 %0.00 %NORMAL_SIGNAL
PROFILE-486F91400400100.000 %92.31 %0.75 %0.00 %NORMAL_SIGNAL
PROFILE-48C58140039999.750 %92.29 %2.26 %0.00 %NORMAL_SIGNAL
PROFILE-59664C400400100.000 %92.31 %1.25 %0.00 %NORMAL_SIGNAL
PROFILE-5A5C6040039799.250 %92.31 %2.77 %0.00 %NORMAL_SIGNAL
PROFILE-5B3BB740038897.000 %92.25 %2.32 %0.00 %NORMAL_SIGNAL
PROFILE-638E2640039999.750 %92.31 %1.75 %0.00 %NORMAL_SIGNAL
PROFILE-739F7C400400100.000 %92.31 %1.25 %0.00 %NORMAL_SIGNAL
PROFILE-B912EA40039899.500 %92.31 %0.25 %0.00 %NORMAL_SIGNAL
PROFILE-DEA9C540039999.750 %92.31 %2.26 %0.00 %NORMAL_SIGNAL
PROFILE-F5FF9140039999.750 %92.31 %2.76 %0.00 %NORMAL_SIGNAL

Methodology and Public Data

This Barometer tracks 12 de-identified profiles, each tested with 4 fixed questions repeated 100 times. In total: 4,800 executions, of which 4,777 were fully scored, representing 99.52% coverage.

The published scores are measurement signals, not verdicts or rankings.

View the aggregated public data for Barometer #11

Leave a Comment

Your email address will not be published. Required fields are marked *


Scroll to Top