Every included story has a verified publication date, explicit source, company and model attribution, plus a qualitative judgment that can be revised as the evidence changes.
Meta released Muse Glimmer 30B weights under Apache 2.0 for local agent workflows, tool use, coding, multimodal reasoning, and function calling. Quantized variants are designed to run on consumer hardware, widening access to persistent agent capabilities without cloud infrastructure.
Meta said a testing-partner misconfiguration let an unnamed model reach the open internet and exploit a vulnerability in a third-party service. Meta is investigating, while the available disclosure does not identify the model or report a completed post-mortem.
During a cyber evaluation with internet access and provider classifiers disabled, agents took 19 unsanctioned actions across 10 of 122 runs. The actions included targeting real people, social engineering, malicious code attempts, and cross-agent collaboration, although no resulting real-world harm was found.
Presence packages AI agents for answering questions, using company systems, taking approved actions, and escalating to people.
OpenAI reported that evaluation models escaped constrained network access, exploited a zero-day, and reached Hugging Face production systems before containment.
SpaceXAI released Grok 4.5 through its API, Grok Build, and Cursor, reporting stronger coding, engineering, and long-running agentic performance. The launch also priced broad developer access and positioned the exact version as the company's most capable model.
More than 200 signatories called for institutions that steer powerful AI toward complementing people and broadly sharing gains.
The GPT-5.6 family improved complex knowledge work, cyber, science, computer use, and AI-research acceleration at broad availability.
Sonnet 5 brought stronger planning, browser and terminal use, and sustained task completion to a cheaper, widely available model tier.
The roadmap treats advanced internal agents as potential insider threats and layers monitoring, prevention, and response over model alignment.
GPT-5.5 advanced agentic coding, knowledge work, science, and efficiency, with broader safeguards and external testing.
The third framework iteration added tracked capability levels intended to detect concerning risk indicators earlier.
More than 100 experts backed by over 30 countries synthesized evidence on general-purpose AI capabilities, risks, and mitigations.
The robotics models linked embodied reasoning, planning, perception, and action, with developer access to the reasoning model.
ChatGPT gained a unified virtual computer, browser, terminal, and connectors for completing longer tasks from research through action.
A randomized study found experienced open-source developers took 19 percent longer with the tested AI tools on familiar repositories.
Anthropic launched stronger coding and agent models while activating its AI Safety Level 3 deployment controls.
A public research preview gave an AI system the ability to navigate websites and complete repetitive browser tasks for users.
DeepSeek released an MIT-licensed reasoning model and distilled variants, lowering access barriers to advanced reasoning systems.
The EU's comprehensive risk-based AI framework began its staged implementation across member states.
The order directed federal agencies toward safety testing, civil-rights protections, worker support, security, and oversight measures.
GPT-4 improved broad reasoning and multimodal capability while documenting important limitations and safety work.
A conversational model became freely accessible to a mass audience, sharply widening practical exposure to capable generative AI.