Tech & AIAI懶人報EP437 - OpenCode、Zeabur、Hugging Face 都出事了:AI 時代,產品還擋得住攻擊嗎?

EP437 - OpenCode、Zeabur、Hugging Face 都出事了:AI 時代,產品還擋得住攻擊嗎?

5.2× channel medianLong-form10:34People & BlogsPublished Oct 19, 2026, 11:27 AM · 3d ago
Views
8,486
+3 views/hour
Views / hour now
3
measured 33m ago
Peak views / hour
188
2d ago
First 24 hours
7.2K
views at 24 h
Engagement
1.0%
likes + comments / views
Likes
79
0.9% of views
Comments
4
0.05% of views
Est. ad revenue
$30–$68
lifetime, $3.5–8 RPM

Checked every 3 h while the video is this age · last check 33m ago · next in 147 min. YouTube itself updates public counts with a short delay.

Avg views / day
2.7K
since upload
vs channel median
5.2×
this video's views ÷ the channel's usual
Rank in channel
#6
of 42 tracked

Views over time

75 readings

Total views Views gained per hour

Description

OpenCode 官方公告(X):https://x.com/opencode/status/2111519637049319818 10 月 17 日,開源 AI 寫程式工具 OpenCode 公布資安事件:入侵者進到一個正在下線的舊系統,用戶存在 OpenCode Console 上的 OpenAI、Anthropic、Google key 可能外洩。這一集從這件事出發,回答一個問題:如果 AI agent 想進來,你的產品擋得住嗎? 擋得住每一次嗎?擋不住。但最近三件真的出事的案子,OpenCode、OpenAI 跟 Hugging Face、Zeabur,攻擊者都不是從最硬的那道門進去的,而是從沒人在顧的地方:快要下線的舊系統、退場中的舊金鑰、外流在網路上的憑證。所以問題要換成三個你控制得了的:你保管了多少別人的鑰匙?你有幾扇快要關掉的門?出事的時候,你多久會知道? ⏱ 章節 00:00 OpenCode 出了什麼事 01:23 AI 攻擊現在走到哪 03:12 OpenAI × Hugging Face 04:25 Zeabur:退場中的舊機制 05:49 三件事放在一起 06:36 你能先做的事 08:22 責任、成本與信任 09:26 整理:擋不住每一次,但可以少輸 📰 事件一手來源 ・The Hacker News|OpenCode Console Breach May Have Exposed OpenAI, Anthropic, and Google API Keys:https://thehackernews.com/2026/10/opencode-console-breach-may-have.html ・Zeabur|2026 年八月資安事件報告:https://zeabur.com/zh-TW/blogs/august-2026-security-incident ・Zeabur|AI Hub 停止服務公告:https://zeabur.com/zh-TW/changelogs/aihub-shutdown ・Zeabur|停止營運公告:https://zeabur.com/zh-TW/blogs/zeabur-is-shutting-down ・OpenAI|Hugging Face 事件說明:https://openai.com/index/hugging-face-model-evaluation-security-incident/ ・OpenAI|事件技術報告(PDF):https://cdn.openai.com/pdf/67869394-cb91-4c12-888c-5cbd85c7814c/OpenAI-Hugging-Face%20Incident-Technical-Report.pdf ・Hugging Face|七月資安事件:https://huggingface.co/blog/security-incident-july-2026 ・Hugging Face|事件技術時間軸:https://huggingface.co/blog/agent-intrusion-technical-timeline 📄 片中數字的來源 ・UK AI Security Institute|Claude Mythos Preview 資安評估:https://www.aisi.gov.uk/blog/our-evaluation-of-claude-mythos-previews-cyber-capabilities ・GPT-5.6 System Card|UK AISI 外部評估:https://deploymentsafety.openai.com/gpt-5-6/external-evaluations-for-alignment-uk-aisi ・Gambit Security|AI agent 攻擊線上零售商:https://gambit.security/blog-posts/autonomous-ai-agents-online-retailers-25-a-company ・Anthropic|Project Glasswing 進度:https://www.anthropic.com/research/glasswing-initial-update ・Anthropic|2026 年 9 月威脅情報報告:https://www.anthropic.com/threat-intelligence-report-september-2026 ・五眼聯盟資安機關聯合聲明:https://www.ncsc.gov.uk/sites/default/files/2026-06/Five-Eyes-cyber-security-agencies-statement-ai-shift.pdf ・歐盟 Cyber Resilience Act:https://eur-lex.europa.eu/eli/reg/2024/2847/oj/eng 🎙 我們之前講過 OpenAI × Hugging Face(EP381):https://www.youtube.com/watch?v=3Y_sSLoMfnc 🧵 參考影片 ・Kurzgesagt – In a Nutshell|AI Just Crossed the Terrifying Line - Now What?:https://youtu.be/ujkD4SxPKOI ・JRE Clips|OpenAI Bots Hacked Hugging Face Without Human Input: Former Researcher Details the Incident:https://youtu.be/hkbv4ILw_II ・Hoog|Hugging Face: Why AI Might Kill Us:https://youtu.be/OO225IfoR3s ・Alberta Tech|The Hugging Face hack is worse than you think:https://youtu.be/_HBLAdk-5uc 片中事實以各家官方公告與報告為準(2026 年 10 月 19 日查核)。OpenCode 的事件還在調查中,後續以官方說明為準。 你的產品現在保管了多少別人的 key?留言告訴我。 歡迎請我喝杯咖啡,幫助我繼續把節目做得更好唷~! 👉 https://buymeacoffee.com/ailanrenbao
Tags
AI agentAI 懶人報AI 資安API keyAnthropicHugging FaceOpenAIOpenCodeZeaburlegacy system蒸餾攻擊資安

Statistics from the YouTube Data API, tracked by FameX. Views per hour, revenue and comparisons are FameX estimates.