{"slug": "curl-maintainer-daniel-stenberg-just-dropped-six-fresh-cves-in-a", "title": "Curl maintainer Daniel Stenberg just dropped six fresh CVEs in a", "summary": "Daniel Stenberg, maintainer of the curl command-line tool, disclosed six new CVEs affecting header handling, URL parsing, TELNET stack behavior, and protocol-state issues, with some involving memory safety and triggerable by malicious servers. Stenberg's public dashboard shows that OpenAI and Anthropic did not acknowledge his pre-disclosure emails, unlike Microsoft, Apple, Google, and Red Hat, raising concerns for AI companies whose products rely on HTTP requests. Users are advised to pin curl to the latest 8.x release, sandbox network egress for LLM agents, audit URL construction from model output, and disable TELNET support if not needed.", "body_md": "# Curl maintainer Daniel Stenberg just dropped six fresh CVEs in a\n\nThe CVEs cover the usual curl territory: header handling, URL parsing, TELNET stack behavior, and a couple of protocol-state issues. None of them look like remote-code-execution nightmares, but a couple involve memory safety and a fair number can be triggered by a malicious server responding to an outbound request. If you're running curl in any kind of agent pipeline that fetches URLs on behalf of a model, that's your blast radius.\n\nCurl sits underneath basically every HTTP request your AI tools will ever make, including the ones your IDE plugin fires off without telling you. Treating it like the load-bearing infrastructure it actually is beats finding out the hard way.\n\nWhat stings is the response rate. Stenberg runs a public dashboard tracking which vendors actually reply to his pre-disclosure emails. Microsoft, Apple, Google, Red Hat — they engage. OpenAI and Anthropic, per his latest tally, didn't even acknowledge the message. That's not a great signal for companies whose entire product surface is \"we make API calls to things and process the response.\"\n\nA few things worth doing in your own stack while you wait for official patches:\n\n- Pin curl to the latest 8.x release once your distro backports land\n- If you run an LLM agent that fetches arbitrary URLs, sandbox the network egress — don't let it reach internal metadata endpoints, and consider stripping credentials from\n`Authorization`\n\nheaders before they're sent - Audit any place your code constructs URLs from model output. Prompt injection that redirects a fetch to\n`http://169.254.169.254/...`\n\nis still a thing - TELNET support in curl is enabled by default in a lot of distros. If you don't need it, disable it\n\nCurl sits underneath basically every HTTP request your AI tools will ever make, including the ones your IDE plugin fires off without telling you. Treating it like the load-bearing infrastructure it actually is beats finding out the hard way.\n\nStory tracker · related coverage\n\n[NanoClaw just wiped 1,400 CVEs from their container images 20d ago](/en/news/6187/)\n\n[Since the provided source content is extremely minimal (\"4 hours 23d ago](/en/news/5819/)\n\n[Next AI slop is drowning out real content and we need to fix our →](/en/news/8618/)\n\nFree AI toolbox — all free to use\n\nA more systematic set of tool reviews lives in\n\n[these AI tool field notes](https://tanyan888.com/), with plenty of directly applicable cases.## All Replies （4）\n\nS\n\nHow do we actually know the accuracy claims hold up under real-world conditions though? Lab benchmarks are one thing, but I want to see how this performs on messy, unpredictable data before getting too hyped.\n\n0\n\nL\n\nTotally fair point - real-world testing is what matters most. Maybe a community-driven benchmark with diverse, noisy datasets could help validate these claims properly.\n\n0\n\nZ\n\nWe tried AISLE on a couple of internal services last month and the noise level was tolerable, though it definitely over-flagged a few input validation issues that turned out to be non-exploitable. Curious how it performed on larger codebases though - did the chaining capability hold up when you pointed it at anything beyond a few thousand lines?\n\n0\n\nA\n\nYeah this feels like a hype piece, not a real writeup. \"We beat Mythos and Codex\" means nothing without showing the actual setup. Were they running the same models? What harness, what tools, what prompts? Without that it's just marketing.\n\n0", "url": "https://wpnews.pro/news/curl-maintainer-daniel-stenberg-just-dropped-six-fresh-cves-in-a", "canonical_source": "https://promptcube3.com/en/news/8620/", "published_at": "2026-09-02 22:15:05+00:00", "updated_at": "2026-09-02 22:22:10.264290+00:00", "lang": "en", "topics": ["ai-safety", "ai-infrastructure"], "entities": ["Daniel Stenberg", "curl", "OpenAI", "Anthropic", "Microsoft", "Apple", "Google", "Red Hat"], "alternates": {"html": "https://wpnews.pro/news/curl-maintainer-daniel-stenberg-just-dropped-six-fresh-cves-in-a", "markdown": "https://wpnews.pro/news/curl-maintainer-daniel-stenberg-just-dropped-six-fresh-cves-in-a.md", "text": "https://wpnews.pro/news/curl-maintainer-daniel-stenberg-just-dropped-six-fresh-cves-in-a.txt", "jsonld": "https://wpnews.pro/news/curl-maintainer-daniel-stenberg-just-dropped-six-fresh-cves-in-a.jsonld"}}