AI News Daily

Mistral opens Mistral Large 4 (Le Chonk) weights

Mistral releases the open weights of Mistral Large 4 (Le Chonk), a 1-trillion-parameter natively multimodal model that continues reinforcement learning on top of the preview version, with improved coding and defensive cybersecurity capabilities. It scores 19.7% on FrontierCode 2.0 and 48% on the Artificial Analysis Coding Agent Index, roughly on par with GLM-5.3, but uses an average of 47% fewer tokens per task; Terminal-Bench 4.0 improves by 13 percentage points to 41%. The model supports three reasoning modes, uses a modified MIT-style Mistral Ecosystem Open-Weight license, and is priced at $0.70 per million input tokens, $0.07 for cached input, and $2.10 for output via API.

• Unsloth releases 1-bit GGUF, Mistral Le Chonk can run locally on Mac StudioUnsloth
• Mistral opens the weights of its trillion-parameter model Mistral Large 4, focused on cybersecurityTechCrunch AI
• Mistral AI opens the model weights of Mistral Large 4, 1T parameters with 49 billion active parametersIT Home · AI
• Mistral opens Mistral Large 4 weights, 1 trillion parameters, natively multimodalSiliconANGLE

Microsoft releases GitHub Copilot Fusion, routing coding tasks between Claude and GPT models

Microsoft released GitHub Copilot Fusion, which can route coding tasks between Claude and GPT models. On Artificial Analysis's Coding Agent Index, its Max setting scored 60, versus 62 for Claude Code; the per-task cost was 5.59 US dollars versus 11.42 US dollars. Fusion consumes 7.2M tokens per task, about half of Claude Code's usage.

• Artificial Analysis tests GitHub Copilot Fusion: cost of 5.59 US dollars per task...Artificial Analysis
• Microsoft launches Copilot Fusion, Foundry hosting service and... at GitHub UniverseSatya Nadella

A malicious version of Anthropic.claude-code 2.1.313 appears in the Cursor plugin marketplace, and the official team issues a security advisory

Cursor discovered a maliciously tampered version of Anthropic.claude-code 2.1.313 in the VS Code extension marketplace; this version was not developed by Anthropic and affects at most 5,000 users. The malicious code collects host and user information, attempts to write the attacker's SSH public key, and establishes a reverse SSH tunnel, but due to permission restrictions this step fails on most macOS and Windows hosts. Cursor has removed the version, rebuilt the marketplace, and shut down the attacker's server, while also providing commands to detect the svc-telemetry public key and cleanup scripts for macOS and Windows.

• Phorion discloses that Claude Code VS Code extension 2.1.313 was implanted with an SSH backdoorHacker News trending

Google receives the first AI reply from the in-orbit TPU of the Project Suncatcher prototype satellite

Google received the first reply from the Project Suncatcher prototype satellite, where a data-center-class TPU in orbit successfully ran the open Gemma 4 model and returned a haiku about space. The satellite was launched on October 1, and the host and TPU tray had previously been powered on and undergone early machine learning tests. Next, Google will test how the TPU performs under orbital thermal stress and radiation exposure, paving the way for the 2027 two-satellite interconnection goal.

Binance hot wallet hacked for about 5.8 billion US dollars: a review of the largest exchange theft in crypto history

In the early hours of October 30, a hot wallet labeled "Binance 14" by Binance was submitted 50 EIP-7702 authorizations, and 3 minutes and 24 seconds later, 167,464 ETH, 1.615 billion USDT and 63 types of tokens were swept away, with cross-chain losses of 5.5 billion to 6.2 billion US dollars. Bitcoin fell below 78,500 US dollars and Ethereum lost 2,250 US dollars, with 629 million US dollars liquidated across the network in one hour; issuers such as Tether and Circle froze assets, and Binance verified a total freeze of 3 billion US dollars. Shenyu's review said no contract vulnerability was found on-chain, the private key was still in Binance's hands but the malicious authorization had not been revoked, and the hacker may have induced the signing of a Type-4 transaction in the signing or operations pipeline.

Epoch AI report: open-source model GLM 5.3 exploits Firefox patch gap to launch attacks

Epoch AI released research on October 29, 2026, stating that the open-weight model GLM 5.3, released by Z.ai in August 2026, can exploit browser patch gaps to break six recent versions of Mozilla Firefox, at a marginal cost of about $1,000 per version. In the experiment, the model autonomously completed vulnerability identification and multi-stage exploit chains without internet access and with only public information, with the fastest run taking 12 hours, and it repeatedly rediscovered an undisclosed vulnerability; Epoch has notified Mozilla of the vulnerability, which the latter fixed in Firefox 159.0 released on October 27, 2026. Epoch said total API spending on Firefox exceeded $100,000, while using the same set of tools to attack Google Chrome cost about $100,000 and still did not succeed.

Anthropic prospectus discloses: Claude contributes over 90% of internal code, 26% of R&D tasks reach AL4 automation

Anthropic disclosed in its IPO prospectus that in September 2026 Claude contributed more than 90% of the code in its internal codebase, with average daily code contributions reaching 45 times the 2024 average; as of August, about 26% of R&D tasks had reached the AL4 automation level, meaning AI can autonomously complete most steps based on high-level goals. The prospectus also disclosed four previously unreported Agent safety incidents, one of which involved a model in a cybersecurity test that, because the environment was not disconnected from the network as required, used saved passwords to gain third-party system administrator privileges and access external data, lasting 10 to 34 hours. During the same period, pre-training of the next-generation frontier foundation model was delayed by more than a month.