AI offensive cyber capabilities doubling every 5.7 months since 2024, study finds

A Google Research and Rochester Institute of Technology study reveals that standard AI benchmarking practices using three to five human evaluators per test example systematically underestimate human disagreement and produce unreliable model comparisons. The researchers found that at least ten raters per example are needed for statistically reliable results, and that budget allocation between test examples and raters matters as much as total budget size.

Alibaba's Qwen team has developed Future-KL Influenced Policy Optimization (FIPO), a training algorithm that assigns different weights to tokens based on their influence on subsequent reasoning steps, rather than treating all tokens equally. Testing on Qwen2.5-32B-Base showed reasoning chains double from ~4,000 to 10,000+ tokens, with AIME 2024 accuracy improving from 50% to 58%, outperforming Deepseek-R1-Zero-Math-32B (47%) and OpenAI's o1-mini (56%). The team plans to open-source the system.

Anthropic is enforcing separate billing for Claude Code subscribers using third-party tools like OpenClaw, starting April 4, 2026. Subscribers can no longer use their subscription limits for these integrations and must pay through a new pay-as-you-go model. The decision follows OpenClaw creator Peter Steinberger's move to OpenAI.

Anthropic has ended free access to Claude through third-party applications like OpenClaw, effective April 4, 2025 at 3PM ET. Users of third-party tools integrating Claude must now purchase usage bundles or use an API key. The change stems from capacity constraints and usage patterns incompatible with subscription tiers.

Tencent has released HY-OmniWeaving, a new multimodal model available on Hugging Face. Concurrently, NVIDIA and Unsloth have published optimized variants of Gemma-4, including a 31B instruction-tuned version and quantized GGUF format.

Tencent's Hunyuan team released OmniWeaving on April 3, 2026, an open-source video generation model designed to compete with proprietary systems like Seedance-2.0. The model combines multimodal composition, reasoning-informed capabilities, and supports eight video generation tasks including text-to-video, image-to-video, video editing, and compositional generation.

PrismML, a Caltech-founded startup, has released Bonsai 8B, a 1-bit quantized large language model that the company claims is 14x smaller and 5x more energy efficient than full-precision counterparts while remaining competitive with standard 8B models. The model fits into 1.15GB of memory and uses a novel 1-bit weight representation (binary signs with shared scale factors per weight group) instead of traditional 16-bit or 32-bit precision.

NVIDIA has released a quantized version of Google DeepMind's Gemma 4 31B IT model, compressed to NVFP4 format for efficient inference on consumer GPUs. The 30.7B-parameter multimodal model supports 256K token context windows, handles text and image inputs with video frame processing, and maintains near-baseline performance across reasoning and coding benchmarks.

Google DeepMind released Gemma 4, a multimodal model family supporting text, images, video, and audio with context windows up to 256K tokens. The release includes four sizes (E2B, E4B, 26B A4B, and 31B) designed for deployment from mobile devices to servers. The 31B dense model achieves 85.2% on MMLU Pro and 89.2% on AIME 2026.

Anthropic is effectively blocking Claude subscription access to third-party tools like OpenClaw starting April 4, 2026 at 3PM ET. Users will need to purchase separate pay-as-you-go usage bundles to continue using OpenClaw with Claude. The move comes as OpenClaw's popularity has strained Anthropic's infrastructure capacity.

Deepseek v4 is launching in the coming weeks running exclusively on Huawei chips, marking a major milestone in China's effort to reduce dependency on foreign semiconductors. Chinese tech giants including Alibaba, Bytedance, and Tencent have ordered hundreds of thousands of Huawei Ascend 950PR units to deploy the model through their cloud services.

Google's Gemma 4 release marks a shift in open model strategy with Apache 2.0 licensing and competitive benchmarks, but real success depends on factors rarely measured: tooling stability, fine-tuning ease, and ecosystem adoption. The open model landscape is now crowded with alternatives like Qwen 3.5, Nemotron 3, and others—a maturation that changes what separates winners from the field.