Tag: proactive measures

  • OpenAI : Estimating worst case frontier risks of open weight LLMs

    Source URL: https://openai.com/index/estimating-worst-case-frontier-risks-of-open-weight-llms Source: OpenAI Title: Estimating worst case frontier risks of open weight LLMs Feedly Summary: In this paper, we study the worst-case frontier risks of releasing gpt-oss. We introduce malicious fine-tuning (MFT), where we attempt to elicit maximum capabilities by fine-tuning gpt-oss to be as capable as possible in two domains: biology and…

  • The Register: Rampant emoji use suggests crypto-stealing NPM package was written by AI

    Source URL: https://www.theregister.com/2025/08/01/emoji_use_ai_malware/ Source: The Register Title: Rampant emoji use suggests crypto-stealing NPM package was written by AI Feedly Summary: Kodane code was either machine-generated or done by a teenager An NPM package packed with cryptocurrency-stealing malware appears to have been largely AI-generated, as evidenced by its liberal use of emojis and other telltale signs.……

  • The Register: OpenAI removes ChatGPT self-doxing option

    Source URL: https://www.theregister.com/2025/08/01/openai_removes_chatgpt_selfdoxing_option/ Source: The Register Title: OpenAI removes ChatGPT self-doxing option Feedly Summary: Checkbox to make chatbot conversations appear in search engines deemed a footgun OpenAI has removed the option to make ChatGPT interactions indexable by search engines to prevent users from unwittingly exposing sensitive information.… AI Summary and Description: Yes Summary: OpenAI’s decision…

  • Embrace The Red: Exfiltrating Your ChatGPT Chat History and Memories With Prompt Injection

    Source URL: https://embracethered.com/blog/posts/2025/chatgpt-chat-history-data-exfiltration/ Source: Embrace The Red Title: Exfiltrating Your ChatGPT Chat History and Memories With Prompt Injection Feedly Summary: In this post we demonstrate how a bypass in OpenAI’s “safe URL” rendering feature allows ChatGPT to send personal information to a third-party server. This can be exploited by an adversary via a prompt injection…

  • Slashdot: Google Confirms It Will Sign the EU AI Code of Practice

    Source URL: https://tech.slashdot.org/story/25/07/30/1916258/google-confirms-it-will-sign-the-eu-ai-code-of-practice?utm_source=rss1.0mainlinkanon&utm_medium=feed Source: Slashdot Title: Google Confirms It Will Sign the EU AI Code of Practice Feedly Summary: AI Summary and Description: Yes Summary: Google has agreed to sign the European Union’s AI Code of Practice, initially opposed for its strict regulations. Despite concerns about potential stifling of innovation, Google aims to influence the…

  • CSA: How to Spot and Stop E-Skimming

    Source URL: https://www.vikingcloud.com/blog/how-to-spot-and-stop-e-skimming-before-it-hijacks-your-customers–and-your-credibility Source: CSA Title: How to Spot and Stop E-Skimming Feedly Summary: AI Summary and Description: Yes Summary: The text explores the growing threat of e-skimming attacks on e-commerce platforms, detailing how cybercriminals exploit JavaScript injections to harvest payment data. It emphasizes the critical need for compliance with PCI DSS v4.x to mitigate…

  • Cloud Blog: Too many threats, too much data, say security and IT leaders. Here’s how to fix that

    Source URL: https://cloud.google.com/blog/products/identity-security/too-many-threats-too-much-data-new-survey-heres-how-to-fix-that/ Source: Cloud Blog Title: Too many threats, too much data, say security and IT leaders. Here’s how to fix that Feedly Summary: An overwhelming volume of threats and data combined with the shortage of skilled threat analysts has left many security and IT leaders believing that their organizations are vulnerable to cyberattacks…