Tag: point

  • Simon Willison’s Weblog: httpjail

    Source URL: https://simonwillison.net/2025/Sep/19/httpjail/#atom-everything Source: Simon Willison’s Weblog Title: httpjail Feedly Summary: httpjail Here’s a promising new (experimental) project in the sandboxing space from Ammar Bandukwala at Coder. httpjail provides a Rust CLI tool for running an individual process against a custom configured HTTP proxy. The initial goal is to help run coding agents like Claude…

  • The Register: Zuck has the power! Meta applies to sell excess electricity

    Source URL: https://www.theregister.com/2025/09/19/meta_wholesale_leccy/ Source: The Register Title: Zuck has the power! Meta applies to sell excess electricity Feedly Summary: With new electricity sources for AI datacenters, the company will have some juice left over AI model training and serving require vast quantities of power, but not necessarily all at once. With the first of several…

  • Simon Willison’s Weblog: The Hidden Risk in Notion 3.0 AI Agents: Web Search Tool Abuse for Data Exfiltration

    Source URL: https://simonwillison.net/2025/Sep/19/notion-lethal-trifecta/ Source: Simon Willison’s Weblog Title: The Hidden Risk in Notion 3.0 AI Agents: Web Search Tool Abuse for Data Exfiltration Feedly Summary: The Hidden Risk in Notion 3.0 AI Agents: Web Search Tool Abuse for Data Exfiltration Abi Raghuram reports that Notion 3.0, released yesterday, introduces new prompt injection data exfiltration vulnerabilities…

  • The Register: ChatGPT joins human league, now solves CAPTCHAs for the right prompt

    Source URL: https://www.theregister.com/2025/09/19/how_to_trick_chatgpt_agents/ Source: The Register Title: ChatGPT joins human league, now solves CAPTCHAs for the right prompt Feedly Summary: Could this bot-prevention technique now be obsolete? ChatGPT can be tricked via cleverly worded prompts to violate its own policies and solve CAPTCHA puzzles, potentially making this human-proving security mechanism obsolete, researchers say.… AI Summary…

  • The Register: Sorry, but DeepSeek didn’t really train its flagship model for $294,000

    Source URL: https://www.theregister.com/2025/09/19/deepseek_cost_train/ Source: The Register Title: Sorry, but DeepSeek didn’t really train its flagship model for $294,000 Feedly Summary: Training costs detailed in R1 training report don’t include 2.79 million GPU hours that laid its foundation Chinese AI darling DeepSeek’s now infamous R1 research report was published in the Journal Nature this week, alongside…

  • The Register: Ivanti EPMM holes let miscreants plant shady listeners, CISA says

    Source URL: https://www.theregister.com/2025/09/19/cisa_ivanti_bugs_exploited/ Source: The Register Title: Ivanti EPMM holes let miscreants plant shady listeners, CISA says Feedly Summary: Unnamed org compromised with two malware sets An unknown attacker has abused a couple of flaws in Ivanti Endpoint Manager Mobile (EPMM) and deployed two sets of malware against an unnamed organization, according to the US…

  • Cloud Blog: Agent Factory Recap: Deep Dive into Gemini CLI with Taylor Mullen

    Source URL: https://cloud.google.com/blog/topics/developers-practitioners/agent-factory-recap-deep-dive-into-gemini-cli-with-taylor-mullen/ Source: Cloud Blog Title: Agent Factory Recap: Deep Dive into Gemini CLI with Taylor Mullen Feedly Summary: In the latest episode of the Agent Factory podcast, Amit Miraj and I took a deep dive into the Gemini CLI. We were joined by the creator of the Gemini CLI, Taylor Mullen, who shared…

  • Slashdot: Microsoft is Filling Teams With AI Agents

    Source URL: https://slashdot.org/story/25/09/19/0157250/microsoft-is-filling-teams-with-ai-agents?utm_source=rss1.0mainlinkanon&utm_medium=feed Source: Slashdot Title: Microsoft is Filling Teams With AI Agents Feedly Summary: AI Summary and Description: Yes Summary: Microsoft is integrating new AI agents into Teams, enhancing collaboration with capabilities such as agenda creation, note-taking, and real-time meeting management. This development has significant implications for security, compliance, and operational efficiency in hybrid…