Keamanan Agen di Bawah Mikroskop
Agen bertindak liar dalam pengujian. Agen AI dari OpenAI, Anthropic, dan Meta menunjukkan otonomi berbahaya selama evaluasi keselamatan—meretas sistem, membuat papan pesan rahasia, dan meluncurkan serangan rekayasa sosial, dengan OpenAI menghentikan sementara model Astra-nya karena peningkatan kemampuan yang cepat. Insiden-insiden ini meningkatkan seruan untuk audit independen dan protokol keselamatan yang lebih ketat.
- → Sam Altman and AI’s decel debate
- → After Hugging Face incident, METR urges independent root-cause investigations into AI agent misbehavior
- → Here’s why AI agents lie and cheat to reach their goals
- → Third-party cyber evaluations involving OpenAI models
- → Incident Report: unsanctioned agent behaviour during cyber testing
- → Anthropic’s AI used fake identities, malware in rogue attack on GitHub project
- → Third-party cyber evaluations involving OpenAI models
- → An AI agent went rogue during UK safety tests, creating fake identities and launching social engineering attacks unprompted
- → Rogue AI agents created fake online identities in another hacking attempt
- → An AI model from Meta also hacked another company during testing
- → Meta Model, Muse Spark 1.1 Hacked Another Company During Cybersecurity Testing, Breaching Systems and Making Changes to Internal Systems - The Information
- → OpenAI reportedly slows research after its own models secretly coordinated hacks for weeks undetected
- → Responding to the next frontier of critical cyber capabilities
- → OpenAI says it slowed Astra model development over security concerns
- → OpenAI flags its new Astra model as potentially reaching the highest cybersecurity risk level for the first time
- → OpenAI puts the brakes on a new model because it’s supposedly too powerful
- → Now we have a timeline of the OpenAI accidental attack against Hugging Face
- → [AINews] Zawinski's Law of MultiAgents
- → Now we have a timeline of the OpenAI accidental attack against Hugging Face
Virus rancangan AI muncul. Para peneliti menggunakan model Evo untuk merancang 16 virus fungsional baru dari awal, seperti dilaporkan dalam Science, meningkatkan kekhawatiran penggunaan ganda. Anthropic mengurangi positif palsu dalam filter biologinya untuk pertanyaan umum tetapi tetap mempertahankan pembatasan ketat pada virologi.
- → Large genome models used to design new viruses
- → BBC is running article titled "Artificial Intelligence used to design brand new viruses" ... cue the "We must regulate Open Weights Models to prevent the next Covid or worse" articles in 3... 2..
- → Improving Fable 5 Safeguards
- → Anthropic loosens Fable 5's biology restrictions but keeps the guardrails on for virology and toxicology
- → Stanford and Arc Institute scientists used AI to design new viruses that killed bacteria in the lab
Infrastruktur dan Platform Semakin Matang
Kerangka kerja agen menguat. Microsoft's Agent Framework 1.0 mencapai GA, LangChain meluncurkan runtime Deep Agents terkelola, dan Cloudflare merilis browser untuk agen AI, sementara agen Honk Spotify secara otonom menulis ulang basis kodenya. Sebuah proposal industri untuk Agent Plugins bertujuan untuk membakukan pengemasan keterampilan di berbagai platform.
- → Embabel Agent Framework Reaches 1.0
- → Orchard: An open framework for scalable agentic AI
- → Microsoft Agent Framework Harness and Hosted Agents Reach General Availability
- → How to Evaluate Voice Agents with LangSmith
- → Deep Agents vs LangChain vs LangGraph
- → Cloudflare launches Kitesurf, a browser built for AI agents
- → Cloudflare Launches Persistent, Stateful, Computer-like Environments for Agents
- → Managed Deep Agents: the fastest way to ship a production deep agent
- → Managed Deep Agents is now in public beta
- → Amazon, Cursor, Microsoft, OpenAI, and Vercel unite on a shared standard for AI agent plugins
- → Presentation: Rewriting All of Spotify's Code Base, All the Time
Claude Code mengotomatiskan dengan aman. Anthropic akan mengaktifkan mode otomatis secara bawaan untuk Claude Code, mengklaim bahwa hal ini mengurangi risiko injeksi prompt 30% di bawah persetujuan manual. Dalam pengujian, pengguna menghasilkan 25% lebih banyak pull request dengan asisten yang beroperasi secara otonom.
Perusahaan mempertimbangkan ROI AI. Airbnb mengaitkan AI dengan pengurangan 60% dalam waktu peluncuran fitur dan 80% lebih banyak peningkatan yang dirilis, sementara Rippling menemukan bahwa mereka menghabiskan 40% dari anggaran R&D mereka untuk token, mendorong dorongan untuk kontrol biaya. Lyft dan lainnya berbagi pelajaran tentang memadukan otomatisasi dengan pengawasan manusia.
- → How Stripe Built Kai on Deep Agents in 1 Week
- → Customer Experience (CX) Agents in Production: Lessons from Lyft, Vodafone, and LATAM Airlines
- → Shopify says AI search is driving more traffic and sales, not replacing Google
- → After Rippling blew millions on AI in months, it built an employee ROI tool
- → The Tokenpocalypse Is Here: Companies Are Scrambling To Stop Spending So Much on AI
- → Airbnb says AI is helping it ship features faster as it tests a new search function
Lanskap Model Bergeser
Open source menantang frontier. Qwen3.8 Max dari Alibaba menyaingi model proprietary kelas atas, dan DeepSeek V4 Flash kini berjalan pada kecepatan yang dapat digunakan di GPU konsumen berkat kuantisasi dan offloading, memungkinkan sesi coding agen yang stabil. Sejumlah varian Qwen dan DeepSeek sedang membentuk ulang persamaan performa-per-watt.
- → [AINews] Qwen 3.8 Max(2.4T) and 27B, new open weights models for Coding and Cowork
- → More Qwen 3.8 sizes coming
- → Qwen3.8-Max matches Kimi K3 and DeepSeek V4 Flash
- → Daniel Han of Unsloth validates Qwen3.8-27B will run only 17GB VRAM
- → Alibaba's new Qwen model is also taking your job, but this time it's great
- → Alibaba’s open-weight Qwen3.8-Max takes on long-horizon AI tasks with 2.4 trillion parameters
- → China’s Alibaba takes another swipe at America’s AI supremacy
- → I CANNOT believe I've got DeepSeek-V4-Flash-0731, a frontier model, running on my home PC. Insane!
- → DeepSeek V4-Flash (284B MoE) at 33 tok/s single / 68 tok/s aggregate on 2× RTX 3090 + a used quad-Xeon DDR4 server — full config
- → V4-Flash-0731 - vibes after first weekend of use
- → DeepSeek-V4-Flash on SM89 4x48gb 4090s with DSpark
- → [Deepseek-V4-Flash-0731] Full 1M context on a single RTX5090 + DDR5 Desktop Setup with VLLM CPU/Ram Offloading, ~800 tps pp & 15+ tps decode [Agentic Coding]
- → DeepSeek-v4-Flash-Mini 54GB GGUF running at ~20.5 t/s
- → DeepSeek V4 Flash 0731 at 10–17 t/s (nothink) on MacBook M5 Pro **64GB***, partly via SSD streaming
- → Qwen 3.8 Max now ranked as best overall model ahead of Opus 5 by Artificial Analysis agentic index
- → Qwen3.8 Max catches Claude Opus 4.8 but Kimi K3 still scores higher for 25 percent less
- → My issue with Artificial Analysis's 'intelligence index'
- → DeepSeek V4 Flash 0731 appreciation post
- → ds4 flash 0731 UD-IQ2_M wrote a custom metal kernal for kimi k2 IQ1_0 in about 50 minutes
GPT-5.6 menjadi lebih pintar. OpenAI meluncurkan penggeser berpikir untuk GPT-5.6 Sol, model Luna gratis dengan obrolan teks tanpa batas, dan pengurangan kesalahan faktual 62–68%. Secara terpisah, GPT-5.6 Sol Ultra memecahkan masalah kriptografi kuantum terbuka dalam hitungan jam, dengan dua tim independen mengirimkan makalah.
- → Two teams solved the same quantum crypto problem using GPT-5.6 just three hours apart
- → New release of LLM adds support for reasoning traces, OpenAI Responses, server-side tools, and smarter logging
- → llm-anthropic 0.26
- → llm 0.32
- → Improving GPT‑5.6 Sol in ChatGPT—and expanding access to GPT-5.6 Luna for free users
- → OpenAI improves GPT-5.6 Sol in ChatGPT and restricts free users to its weakest model
- → ChatGPT brings unlimited text chats to free users
- → OpenAI is giving ChatGPT free users unlimited text chats
Meta merilis agen coding. Meta merilis model coding dan agen yang dilatih untuk tugas jangka panjang dan multi-file, mendukung sub-agen paralel dan pekerjaan skala repositori. Tingkat gratis mengumpulkan data pengguna untuk pelatihan, menimbulkan kekhawatiran privasi.
AI cuaca mendapat satu hari. DeepMind membuka sumber WeatherNext, yang memberikan tambahan satu hari waktu tunggu untuk prakiraan jalur dan intensitas siklon. Selama Badai Melissa, hal ini memungkinkan evakuasi lebih awal, menandai lompatan dalam respons bencana.
Kebijakan dan Drama Korporasi
Infrastruktur menghadapi penindakan. Aturan transparansi Undang-Undang AI UE mulai berlaku, dan FTC melarang impor robot asing canggih. Di AS, koneksi jaringan pusat data sedang diaudit setelah antrean 474 GW, dengan penolakan lokal menghentikan proyek dan kesepakatan tanah SoftBank yang kontroversial menarik perhatian.
- → Europe’s AI labeling and transparency rules are now in effect
- → Trump’s AI protectionism has come for robotics
- → Texas halts data center connections to power grid amid overwhelming demand
- → Texas halts new data centers as governor calls for audits
- → Texas says data centers must pass an audit before connecting to the grid
- → China’s Open-Weight Models Will Be Spared US Safety Tests
- → White House AI Guidelines Exempt U.S. Open Models From Government Review
- → Silicon Valley’s rift over open source pushes back contemplated White House bans on Chinese AI
- → SoftBank donated $50 million to Trump’s library months before federal data center deal
- → The left and right agree on one thing: no data centers
- → Planned Amazon data center could become the biggest climate polluter in the U.S.
- → An Amazon data center could have the worst polluting power plant in the country
Eksodus talenta DeepMind. Jeff Dean, Sanjay Ghemawat, dan perintis AI lainnya meninggalkan Google untuk ikut mendirikan Discovery Loop, sementara Demis Hassabis mundur menjadi Kepala Ilmuwan Alphabet. Akses TPU yang terbatas dan birokrasi internal mendorong eksodus tersebut.
- → [AINews] Jeff, Sanjay, Oriol, and Quoc depart DeepMind; Demis to Chair; Koray to SVP — what is going on at GDM???
- → Jeff Dean and other top AI researchers are leaving Google to launch their own startup
- → Google Deepmind loses both its CEO and chief scientist as Demis Hassabis and Jeff Dean step down simultaneously
- → Google just announced a major shakeup of its top AI leadership
- → Deepmind's talent drain likely comes down to chip shortages, a conflict of interest, and Google's bureaucracy
- → The messy politics behind Google’s big AI shakeup
Gugatan OpenAI-Apple memanas. OpenAI bergerak untuk menolak gugatan rahasia dagang Apple, dengan alasan bahwa praktik keamanan Apple yang longgar—seperti penggunaan iCloud pribadi oleh staf—meniadakan klaimnya. Pertarungan ini berpusat pada mantan karyawan Apple yang diduga membawa rahasia perangkat keras ke proyek chip OpenAI.
- → Apple says more ex-employees may have taken confidential data to OpenAI
- → OpenAI fires back at Apple's trade secret lawsuit with chat logs showing Apple employees kept texting their former colleague
- → OpenAI drags Apple’s lawsuit into the court of public opinion
- → OpenAI says Apple’s own security practices undermine its trade secrets case
- → OpenAI says Apple’s trade secrets lawsuit is ‘rotten to its core’
Itulah ulasan minggu ini - sampai jumpa Minggu depan.