#Safety/Alignment

AI…the Index. Compared to the publicly available GPT-6 Sol, the Daybreak Blue model has the largest gains on CyberGym-E2E, which is the benchmark where we observe the most safety refusals

AIArtificial Analysis added trusted-access models to its Cyber Index, and GPT-6 Sol (Daybreak Blue, max) now ranks first. The model is available only through OpenAI's Daybreak program and records no safety blocks across the Index. Its overall score is 32 points higher than the publicly available GPT-6 Sol (max), at a cost of $1.77 per task versus $11.67 for Grok 4.7 (xhigh).

AIThe team released Carbon-A, an open model that finds genes directly in DNA, along with a database of 566.34 million candidate genes across 22,617 species. The model reads genomes without needing a close relative, and wet-lab validation in cats, chickens, and arabidopsis is cited, with 239 genes found missing from reference annotations of common species. The authors say the model marks gene locations but does not design DNA or predict gene function.

AIArtificial Analysis is adding trusted-access models to its Cyber Index, starting with GPT-6 Sol (Daybreak Blue, max), which is available only through OpenAI's Daybreak program. The model hits no safety blocks across the Index and scores 32 points higher overall than the publicly available GPT-6 Sol (max), with its largest gains on CyberGym-E2E.

Why it matters: The source shows how safety refusals shape cyber benchmark scores, with the trusted-access model's gains concentrated on CyberGym-E2E, useful for comparing guarded and unguarded models.

AIBuilt by vLLM Semantic Router and KR Labs. Four sizes, 0.3B to 9B. Apache-2.0.

AIMistral says its ML4 model reaches state-of-the-art performance among open models across a wide range of capabilities, and outperforms the best models in visual grounding, legal, and spreadsheet manipulation. The post reports ML4 ranks among the best on the AA Cyber Index, scoring 82% on vulnerability reproduction and patching and 93% on Cybench. It argues that self-hosted, auditable open models are the best defense option for enterprises today, and that they do not refuse to help.

AIGoogle is sharing Gemini 4 Argon first with trusted defenders in its Fairwind Program, with frontier capabilities in coding, knowledge work, and cyber security defense. The model is also being rolled out to the US government, and the company plans broader availability as testing progresses and safeguards allow.

AIIt delivers frontier performance across the board on complex software tasks. Thousands of Googlers have already been using it in Antigravity internally and we’re excited by the feedback. It’s rolling out first to a set of trusted cyber defenders in our Fairwind Program with broader availability coming as soon as possible.

AI…cyber defenders through our Fairwind Program today. We’re going to make it available as soon as we can and as safely as we can. So hold tight, lots more coming, and you’re going to see us iterating rapidly.

AIGoogle DeepMind announced Gemini 4 Argon, rolling out first to trusted cyber defenders through its Fairwind Program. Argon will launch at an introductory price of $2 per million input tokens and $10 per million output tokens, with output limits raised to 1M tokens. The post cites a 77.9% score on DeepSWE v1.1 and 91.7% on LVBench, and says broad availability will follow safeguard testing.

Why it matters: The post pairs Argon's benchmark claims with the phased release, pricing, and safeguard details, helping readers weigh its frontier-level capabilities against its access limits.

AIGoogle announced Gemini 4 Argon, a new frontier model rolling out first to trusted cyber defenders through its Fairwind Program. The model's output limit rises to 1M tokens from 64K, and its introductory API price is $2 per million input tokens and $10 per million output tokens. Google says broader availability to developers, enterprises, and consumers will follow after more testing of guardrails.

Why it matters: The post pairs benchmark claims with a phased access plan, pricing, and safety measures, which helps readers judge how quickly Argon may reach developers.

AIIt’s more transparent about its limitations and more reliable at respecting user intent and safety constraints.

AIGoogle DeepMind's latest Gemini Audio models let developers design new vocal personas from natural language prompts, directing pacing, back channeling, and dialect shifts line by line. Developers can also recreate consistent adult voice profiles from a 30-second audio sample, with built-in consent verification, SynthID watermarking, and C2PA credentials.

AIxAI released Grok 4.7, which it calls its most capable model for coding and knowledge work, built on a larger base model than Grok 4.6 and trained with a longer reinforcement learning run. It is priced from $2 per million input tokens and $6 per million output tokens, the same as Grok 4.6, and is available in Cursor, Grok Build, and the Grok API. xAI reports gains on CursorBench 4.0 (46.3%) and AA Briefcase v1.1 (1,657) over Grok 4.6, and says it posts the strongest safety results it has tested on refusals and jailbreak resistance.

Why it matters: The release pairs a new base model with benchmark tables against named rivals and pricing, letting readers compare its coding and office-work gains against Grok 4.6 and frontier models.

AIinclusionAI released Step-3.7-Flash-singprobe, an 8.13M-parameter probe that reuses Step-3.7-Flash hidden states to score query intent, response unsafety, and hallucination risk at every generated token. The probe adds less than 0.5% decode-time overhead and reports 0.9858 R-AUC and 0.9295 T-AUC on streaming safety benchmarks. It is supported through SGLang and vLLM integration branches and loads from Hugging Face by checkpoint ID.

AIinclusionAI has released Qwen3.8-27B-singprobe, a 10.1M-parameter intrinsic streaming guardrail that reuses Qwen3.8-27B hidden states to score query intent, response unsafety, and hallucination risk at every token. The probe adds less than 0.5% decode-time overhead and reports a 0.03% benign-response false-positive rate averaged across five datasets. It is supported through SGLang and vLLM integration branches, with training code available at inclusionAI/SingProbe.

AIinclusionAI has released Qwen3.5-397B-A17B-singprobe, an intrinsic streaming guardrail built on Qwen/Qwen3.5-397B-A17B that scores query intent, response unsafety, and hallucination risk at every generated token using the base model's hidden states. The probe has 8.13M parameters, taps layers 18, 38, and 58, and adds less than 0.5% decode-time overhead. Training code is available at inclusionAI/SingProbe, and the probe runs through SGLang or vLLM integration branches.

AIinclusionAI released SingProbe, a 5.8M-parameter intrinsic guardrail built on openai/gpt-oss-120b that scores query intent, response unsafety, and hallucination risk at every token. It reuses the base model's hidden states, adding less than 0.5% decode-time overhead, and reports a 0.06% benign-response false-positive rate. The probe is available on Hugging Face and supported through SGLang and vLLM integrations.

AIOpenAI researcher Mark Chen announced GPT-6 Astra, which he described as the company's most capable and aligned model yet. He said it can build and test software, work across apps on a computer, and help with open scientific problems. The post also highlights improved computer use compared with Operator and stronger monitoring that can stop potentially unauthorized agent actions.

Why it matters: The post links a named model release to specific capabilities like computer use and aligned agent behavior, giving readers concrete claims to check against the model.

AIFor example, in evaluations on real Chrome security bugs, 3.8 Flash Cyber produced 2.6x more correct patches for vulnerabilities than larger models. Because of its advanced capabilities, we're giving Government agencies & cybersecurity partners access through the Fairwind program, read more:

AIGoogle introduces Gemini 3.8 Flash Cyber, which it describes as its most capable cybersecurity model. The company reports 86.2% on CyberGym, 47.2% on CWE-Bench for patching, and a 70%+ success rate in discovering vulnerabilities across 20 programming languages on its internal benchmark. Google says the model offers frontier-level performance at Flash-level speed and pricing.

AIGoogle launches Gemini 3.8 Flash Cyber and Gemini 3.8 Flash. The post describes Flash Cyber as its most capable cybersecurity model for finding and fixing vulnerabilities, placing it on the Pareto frontier for patching on CWE-Bench. Flash Cyber is available to trusted defenders through the new Fairwind Program.

AIZ.ai has made GLM-5.3 open-weight, so users can download, run, and customize its weights. The company describes it as its most capable model for agentic coding and cyber defense, with weights on Hugging Face and details in a tech blog.

AIZ.ai has released GLM-5.3-Flash, a model with native visual capabilities that observe interfaces, rendering results, and interaction feedback across code, browsers, and GUIs. It uses a hybrid linear and sparse attention architecture with 320B total parameters and 18B activated, which the company says significantly reduces compute and KV-cache requirements. The release notes also describe support for office document and financial research workflows.

Why it matters: The release notes give GLM-5.3-Flash's architecture, parameter counts, and cybersecurity findings, which make the model's scope concrete for comparison with earlier GLM releases.

AIZ.ai says GLM-5.3 is available now through GLM Coding Plan and ZCode. API access and open weights will follow in stages after rigorous safety evaluations.

AIGoogle introduces Gemini 3.6 Flash as its workhorse model, with better coding, knowledge work, and multimodal performance while reducing token usage. It also launches Gemini 3.5 Flash-Lite, described as the fastest and most cost-effective 3.5-class model for high-throughput applications, and 3.5 Flash Cyber, a version of 3.5 Flash fine-tuned to find and fix cybersecurity vulnerabilities.

Why it matters: The post lists three distinct models, each aimed at a different job, so readers can map which one fits coding, high-volume, or security workloads.

AIMistral AI released Shieldstral-1.0-3B, a 3B-parameter multimodal safety classifier that judges content against natural-language policies and outputs a continuous safety score. It moderates text, image, and text-plus-image content in a single forward pass and can be retargeted to new policies at inference time without retraining. The Apache 2.0 open-weight model is built on Ministral-3-3B-Base-2512 and trained on sequences up to 32k tokens.

AIEugene Yan relays two evaluations of Mythos: UK AISI reports it completed a 32-step network attack, estimated at about 20 expert hours, in 6 of 10 tries and was the first model to solve its end-to-end cyber ranges. XBOW's evaluation describes its performance as token-for-token and unprecedented in precision. The post links both AISI and XBOW blog posts for details.