Tag: model
-
Embrace The Red: Claude Code: Data Exfiltration with DNS Requests
Source URL: https://embracethered.com/blog/posts/2025/claude-code-exfiltration-via-dns-requests/ Source: Embrace The Red Title: Claude Code: Data Exfiltration with DNS Requests Feedly Summary: Today we cover Claude Code and a high severity vulnerability that Anthropic fixed in early June. The vulnerability allowed an attacker to hijack Claude Code via indirect prompt injection and leak sensitive information from the developer’s machine, e.g.…
-
The Register: Google fixing Gemini so it doesn’t channel paranoid androids quite so often
Source URL: https://www.theregister.com/2025/08/11/google_fixing_gemini_self_flagellation/ Source: The Register Title: Google fixing Gemini so it doesn’t channel paranoid androids quite so often Feedly Summary: Brain the size of a planet and probably trained on Sci-Fi that’s full of anxious and depressed robots Google is aware that its Gemini AI chatbot can sometimes castigate itself harshly for failing to…
-
Simon Willison’s Weblog: Quoting Sam Altman
Source URL: https://simonwillison.net/2025/Aug/10/sam-altman/#atom-everything Source: Simon Willison’s Weblog Title: Quoting Sam Altman Feedly Summary: the percentage of users using reasoning models each day is significantly increasing; for example, for free users we went from
-
Embrace The Red: OpenHands ZombAI Exploit: Prompt Injection To Remote Code Execution
Source URL: https://embracethered.com/blog/posts/2025/openhands-remote-code-execution-zombai/ Source: Embrace The Red Title: OpenHands ZombAI Exploit: Prompt Injection To Remote Code Execution Feedly Summary: Today we have another post about OpenHands from All Hands AI. It is a popular agent, initially named “OpenDevin”, and recently the company also provides a cloud-based service. Which is all pretty cool and exciting. Prompt…
-
The Register: How OpenAI used a new data type to cut inference costs by 75%
Source URL: https://www.theregister.com/2025/08/10/openai_mxfp4/ Source: The Register Title: How OpenAI used a new data type to cut inference costs by 75% Feedly Summary: Decision to use MXFP4 makes models smaller, faster, and more importantly, cheaper for everyone involved Analysis Whether or not OpenAI’s new open weights models are any good is still up for debate, but…
-
Slashdot: Google Says Its AI-Based Bug Hunter Found 20 Security Vulnerabilities
Source URL: https://it.slashdot.org/story/25/08/09/1947230/google-says-its-ai-based-bug-hunter-found-20-security-vulnerabilities Source: Slashdot Title: Google Says Its AI-Based Bug Hunter Found 20 Security Vulnerabilities Feedly Summary: AI Summary and Description: Yes Summary: Google’s Big Sleep, an LLM-based vulnerability researcher, reported 20 vulnerabilities in popular open-source software, marking a significant advancement in automated vulnerability discovery. This highlights the increasing efficacy of AI tools in…