Tag: Microsoft Copilot
-
Hacker News: Microsoft Copilot continues to expose private GitHub repositories
Source URL: https://www.developer-tech.com/news/microsoft-copilot-continues-to-expose-private-github-repositories/ Source: Hacker News Title: Microsoft Copilot continues to expose private GitHub repositories Feedly Summary: Comments AI Summary and Description: Yes **Summary:** The investigation by Lasso into claims about ChatGPT accessing private GitHub repositories highlighted critical concerns regarding data privacy and the phenomenon known as “Zombie Data.” This situation underscores the risks of…
-
Slashdot: Thousands of Exposed GitHub Repositories, Now Private, Can Still Be Accessed Through Copilot
Source URL: https://yro.slashdot.org/story/25/02/27/2129241/thousands-of-exposed-github-repositories-now-private-can-still-be-accessed-through-copilot?utm_source=rss1.0mainlinkanon&utm_medium=feed Source: Slashdot Title: Thousands of Exposed GitHub Repositories, Now Private, Can Still Be Accessed Through Copilot Feedly Summary: AI Summary and Description: Yes Summary: This text highlights significant security concerns raised by researchers regarding potential data exposure through generative AI tools like Microsoft Copilot. It underscores the persistence of data that can…
-
Hacker News: Exposed GitHub repos, now private, can be accessed through Copilot
Source URL: https://techcrunch.com/2025/02/26/thousands-of-exposed-github-repos-now-private-can-still-be-accessed-through-copilot/ Source: Hacker News Title: Exposed GitHub repos, now private, can be accessed through Copilot Feedly Summary: Comments AI Summary and Description: Yes Summary: The text discusses the risks associated with data exposure in generative AI systems, particularly focusing on Microsoft Copilot’s ability to access previously public data from GitHub repositories, even after…
-
The Register: What better place to inject OpenAI’s o1 than Los Alamos national lab, right?
Source URL: https://www.theregister.com/2025/01/30/openai_los_alamos_national_lab/ Source: The Register Title: What better place to inject OpenAI’s o1 than Los Alamos national lab, right? Feedly Summary: Tackling disease, tick. High-energy physics, tick. Nuclear weapon security, also tick OpenAI has announced another deal with Uncle Sam, this time to get its very latest models in the hands of US government…
-
AlgorithmWatch: Large language models continue to be unreliable concerning elections
Source URL: https://algorithmwatch.org/en/llms_state_elections/ Source: AlgorithmWatch Title: Large language models continue to be unreliable concerning elections Feedly Summary: Large language models continue to be unreliable for election information. Our research was able to substantially improve the reliability of safeguards in the Microsoft Copilot chatbot against election misinformation in German. However barriers to data access greatly restricted…
-
Simon Willison’s Weblog: LLM Flowbreaking
Source URL: https://simonwillison.net/2024/Nov/29/llm-flowbreaking/#atom-everything Source: Simon Willison’s Weblog Title: LLM Flowbreaking Feedly Summary: LLM Flowbreaking Gadi Evron from Knostic: We propose that LLM Flowbreaking, following jailbreaking and prompt injection, joins as the third on the growing list of LLM attack types. Flowbreaking is less about whether prompt or response guardrails can be bypassed, and more about…