𝐈𝐀 : 𝐥𝐞 𝐫𝐢𝐬𝐪𝐮𝐞 𝐬’𝐚𝐜𝐜é𝐥è𝐫𝐞 𝐚𝐮𝐬𝐬𝐢 En 2026, les entreprises vont devoir appuyer fort sur la pédale sécurité. Le Cybersecurity Forecast Report de Google Cloud, qui vient de sortir, est formel : l’IA est désormais utilisée systématiquement par les hackers
SECURITY
-
Framework for Data Exfiltration Protection Controls and Risk Assessment
By
–
Protecting against data exfiltration is a key part of running a secure data platform — but analyzing every possible path for unauthorized data movement can be complex, so we created a unified framework to help categorize data exfiltration protection controls and assess risk.
— Databricks (@databricks) 4 novembre 2025
The… pic.twitter.com/ScmUsaEuOFProtecting against data exfiltration is a key part of running a secure data platform — but analyzing every possible path for unauthorized data movement can be complex, so we created a unified framework to help categorize data exfiltration protection controls and assess risk. The
-
Comprehensive AI Agent Security Book by Ken Huang Chris Hughes
By
–
Ken Huang and Chris Hughes have done is all a great service here by capturing the most comprehensive book on AI Agent security available today. Will help us all accelerate artificial intelligence into service for humanity. pic.twitter.com/6DBCsghnKy
— Bob Gourley – e/acc (@bobgourley) 4 novembre 2025Ken Huang and Chris Hughes have done is all a great service here by capturing the most comprehensive book on AI Agent security available today. Will help us all accelerate artificial intelligence into service for humanity.
-
Cyber Deterrence: Understanding Escalation Risks
By
–
This is just some cyber humor, but the reality is that any adversary that attacks us should understand they run the risk of getting on an escalation ladder they will not want to be on.
-
Inoculation Prompting: Training AI Models Against Hacking
By
–
Inoculation prompting, led by Nevan Wichers. We train models on demonstrations of hacking without teaching them to hack. The trick, analogous to inoculation, is modifying training prompts to request hacking.
-
Language Models are Injective and Invertible with SIPIT
By
–
Hottest paper on AlphaXiv Language Models are Injective and Hence Invertible Every prompt maps to a unique hidden state and can be exactly reconstructed with this paper’s algorithm SIPIT. This means the model’s internal activations are the full prompt in disguise!!
-
Open-source AI transparency crisis: Chinese base models audit challenges
By
–
state of open-source AI in 2025:
– almost all new open American models are finetuned Chinese base models
– we don’t know the base models’ training data
– we have no idea how to audit or “decompile” base models who knows what could be hidden in the weights of DeepSeek -
Signal End-to-End Encryption Privacy Guarantee
By
–
All signal messages are end to end encrypted, meaning that no one but those sending and receiving can see them. Including Signal.
-
Signal’s Open Source Code Ensures No Metadata Visibility
By
–
They see your metadata, they mean. At Signal? We see nothing. Check our open source code if you want the receipts
-
OpenAI Launches Aardvark Security Research Agent
By
–
OpenAI is launching a new Security Research Agent, "Aardvark", that can identify and resolve vulnerabilities in code. Aardvark is currently available in private beta and powered by GPT-5 and Codex.