Skip to main content
Scour
Discover
Docs
Login
Sign Up
You are offline. Trying to reconnect...
Copied to clipboard
Unable to share or copy to clipboard
The AI Security Institute (AISI)
aisi.gov.uk
AI Security Institute
·
19h
19 hours ago
RealityTest: How People Probe AI Identity and Whether Models Disclose It
Love
Like
Not for me
Save
See related topics
Feeds
Share
Report
Spam
Misleading
Harmful Content
Block Domain
Actions for RealityTest: How People Probe AI Identity and Whether Models Disclose It
AI Security Institute
·
19h
19 hours ago
Conference on frontier AI safety frameworks | AISI Work
Love
Like
Not for me
Save
See related topics
Feeds
Share
Report
Spam
Misleading
Harmful Content
Block Domain
Actions for Conference on frontier AI safety frameworks | AISI Work
AI Security Institute
·
19h
19 hours ago
UKAISI at NeurIPS 2025 | AISI Work
Covers
2 stories
See all stories this covers
including
Deep Ignorance: Filtering Pretraining Data Builds Tamper-Resistant Safeguards into Open-Weight LLMs
Love
Like
Not for me
Save
See related topics
Feeds
Share
Report
Spam
Misleading
Harmful Content
Block Domain
Actions for UKAISI at NeurIPS 2025 | AISI Work
AI Security Institute
·
19h
19 hours ago
Announcing Inspect Evals | AISI Work
Love
Like
Not for me
Save
See related topics
Feeds
Share
Report
Spam
Misleading
Harmful Content
Block Domain
Actions for Announcing Inspect Evals | AISI Work
AI Security Institute
·
19h
19 hours ago
International consensus and open questions in AI evaluations | AISI Work
Covers
NeurIPS Tightens Sanctions Compliance
Covered by
Tech Policy Press
Love
Like
Not for me
Save
See related topics
Feeds
Share
Report
Spam
Misleading
Harmful Content
Block Domain
Actions for International consensus and open questions in AI evaluations | AISI Work
AI Security Institute
·
19h
19 hours ago
Security challenges in AI agent deployment: Insights from a large scale public competition
Love
Like
Not for me
Save
See related topics
Feeds
Share
Report
Spam
Misleading
Harmful Content
Block Domain
Actions for Security challenges in AI agent deployment: Insights from a large scale public competition
AI Security Institute
·
19h
19 hours ago
Making safeguard evaluations actionable | AISI Work
Love
Like
Not for me
Save
See related topics
Feeds
Share
Report
Spam
Misleading
Harmful Content
Block Domain
Actions for Making safeguard evaluations actionable | AISI Work
AI Security Institute
·
19h
19 hours ago
Strengthening AI resilience | AISI Work
Love
Like
Not for me
Save
See related topics
Feeds
Share
Report
Spam
Misleading
Harmful Content
Block Domain
Actions for Strengthening AI resilience | AISI Work
AI Security Institute
·
19h
19 hours ago
More compute, more capability: Why AI agent evaluations need to account for test-time compute | AISI Work
Covers
2 stories
See all stories this covers
including
Measuring AI Ability to Complete Long Tasks
Covered by
4 sources
See all sources covering this story
including
METR
,
echohive.ai
Love
Like
Not for me
Save
See related topics
Feeds
Share
Report
Spam
Misleading
Harmful Content
Block Domain
Actions for More compute, more capability: Why AI agent evaluations need to account for test-time compute | AISI Work
AI Security Institute
·
19h
19 hours ago
Advancing the field of systemic AI safety: grants open | AISI Work
Love
Like
Not for me
Save
See related topics
Feeds
Share
Report
Spam
Misleading
Harmful Content
Block Domain
Actions for Advancing the field of systemic AI safety: grants open | AISI Work
AI Security Institute
·
19h
19 hours ago
Alignment Pretraining: AI Discourse Causes Self-Fulfilling (Mis)alignment
Love
Like
Not for me
Save
See related topics
Feeds
Share
Report
Spam
Misleading
Harmful Content
Block Domain
Actions for Alignment Pretraining: AI Discourse Causes Self-Fulfilling (Mis)alignment
AI Security Institute
·
19h
19 hours ago
A multi-turn framework for evaluating AI misuse in fraud and cybercrime scenarios
Love
Like
Not for me
Save
See related topics
Feeds
Share
Report
Spam
Misleading
Harmful Content
Block Domain
Actions for A multi-turn framework for evaluating AI misuse in fraud and cybercrime scenarios
AI Security Institute
·
19h
19 hours ago
How to evaluate control measures for AI agents? | AISI Work
Covers
2 stories
See all stories this covers
including
Alignment faking in large language models
Love
Like
Not for me
Save
See related topics
Feeds
Share
Report
Spam
Misleading
Harmful Content
Block Domain
Actions for How to evaluate control measures for AI agents? | AISI Work
AI Security Institute
·
19h
19 hours ago
Pre-Deployment evaluation of OpenAI’s o1 model | AISI Work
Love
Like
Not for me
Save
See related topics
Feeds
Share
Report
Spam
Misleading
Harmful Content
Block Domain
Actions for Pre-Deployment evaluation of OpenAI’s o1 model | AISI Work
AI Security Institute
·
19h
19 hours ago
How our Control Red Team is stress-testing frontier monitors | AISI Work
Covers
3 stories
See all stories this covers
including
Claude Code auto mode: a safer way to skip permissions
Love
Like
Not for me
Save
See related topics
Feeds
Share
Report
Spam
Misleading
Harmful Content
Block Domain
Actions for How our Control Red Team is stress-testing frontier monitors | AISI Work
AI Security Institute
·
19h
19 hours ago
Finding Cloud Misconfigurations with Frontier AI: A Case Study | AISI Work
Covers
3 stories
See all stories this covers
including
Inspect
Covered by
echohive.ai
Love
Like
Not for me
Save
See related topics
Feeds
Share
Report
Spam
Misleading
Harmful Content
Block Domain
Actions for Finding Cloud Misconfigurations with Frontier AI: A Case Study | AISI Work
AI Security Institute
·
19h
19 hours ago
Safety cases at AISI | AISI Work
Love
Like
Not for me
Save
See related topics
Feeds
Share
Report
Spam
Misleading
Harmful Content
Block Domain
Actions for Safety cases at AISI | AISI Work
AI Security Institute
·
19h
19 hours ago
Async control: Stress-testing asynchronous control measures for LLM agents
Love
Like
Not for me
Save
See related topics
Feeds
Share
Report
Spam
Misleading
Harmful Content
Block Domain
Actions for Async control: Stress-testing asynchronous control measures for LLM agents
AI Security Institute
·
19h
19 hours ago
An evaluation framework for AI misuse in fraud and cybercrime | AISI Work
Covers
Detecting and countering misuse of AI: August 2025
Love
Like
Not for me
Save
See related topics
Feeds
Share
Report
Spam
Misleading
Harmful Content
Block Domain
Actions for An evaluation framework for AI misuse in fraud and cybercrime | AISI Work
AI Security Institute
·
19h
19 hours ago
How can safety cases be used to help with frontier AI safety? | AISI Work
Covers
AI Sandbagging: Language Models can Strategically Underperform on Evaluations
Love
Like
Not for me
Save
See related topics
Feeds
Share
Report
Spam
Misleading
Harmful Content
Block Domain
Actions for How can safety cases be used to help with frontier AI safety? | AISI Work
« Page 3
·
Page 5 »
Log in to enable infinite scrolling
Keyboard Shortcuts
Navigation
Next / previous post
j
/
k
Open post
o
or
Enter
Preview post
v
Post Actions
Love post
a
Like post
l
Dislike post
d
Undo reaction
u
Save / unsave
s
Recommendations
Add interest / feed
Enter
Not interested
x
Go to
Home
g
h
Interests
g
i
Feeds
g
f
Likes
g
l
History
g
y
Changelog
g
c
Settings
g
s
Discover
g
b
Search
/
Pagination
Next page
n
Previous page
p
General
Show this help
?
Submit feedback
!
Close modal / unfocus
Esc
Press
?
anytime to show this help
Like
Save
Not for me
Report