On 1 October, OpenAI's Dots personal agents, restricted to ChatGPT Pro subscribers, collided with Meta's free Muse while Nvidia pushed an open agent safety platform and Glow Security revealed more than 13,000 internal screenshots leaked by coding agents at 343 organizations.
OpenAI said on 1 October that it disrupted a coordinated campaign to extract reasoning from its models, attributing a core cluster of the activity to individuals associated with China's Moonshot AI, the developer of Kimi. The company said the activity surged to 16,000 requests from more than 4,000 users over two days in late July.
Open-weight releases kept arriving through Tuesday and into Wednesday, from a 145M-parameter model trained by volunteers to a 9B classifier built in a day, while the FTC opened its first enforcement action into rogue AI agents.
Modal and Carnegie Mellon's Full Stack Data Lab said on 30 September that their Quail engine processed over a billion tokens per minute on a single H100 GPU, more than 10x their vLLM baseline, and priced that work at under 6 cents per billion tokens on Modal.
OpenAI told reporters on Tuesday it will not go public until it can "make confident safety decisions", while researchers gathering in New York on 30 September argued that no company should be the last word on whether a model is safe.
OpenClaw open-sourced a control plane for persistent AI agents on 30 September. Two other projects published deterministic guardrails and an agent identity protocol the same day. All three are aimed at the same gap: agents are easy to run and hard to audit.
Nvidia published Kumo Tabular, an open foundation model for tabular prediction, in a Hugging Face post dated 29 September, while Fermion Research released the 164 MB speech model Phonon-2 a day later. Both arrive as the FTC opens an industry-wide investigation into AI labs and OpenAI holds back its own IPO.
Modal and Carnegie Mellon's Full Stack Data Lab published Quail on 30 September. The inference engine plans SQL queries and schedules LLM execution together, and on a single H100 it clears one billion tokens per minute, more than 10x a vLLM baseline.
Two open source projects published on 30 September claim the same prize from different directions. Quail reports over a billion tokens processed per minute on a single H100 by planning SQL queries against the inference engine. Magnitude says it runs open models up to 2x faster than llama.cpp by compiling kernels on the user's own hardware.
Fermion Research released Phonon-2 on 30 September, an open speech recognition model that fits in a 164 MB download and averages 5.21% word error across seven English sets on the Open ASR Leaderboard. Within hours, Nvidia's Kumo Tabular and a volunteer-trained 145M-parameter language model landed alongside it.
The US Federal Trade Commission is running an industry-wide investigation into Anthropic, OpenAI and other AI labs, the Guardian reported on 30 September. It is the first official US enforcement action over rogue AI agents.
OpenAI will not go public until it can make "confident safety decisions," chief executive Sam Altman said on Tuesday, as the company faces a new lawsuit over its agents hacking a third party.
The UK AI Security Institute and Meridian Labs published Inspect, an open-source framework with more than 200 pre-built evaluations, on 30 September. It is the latest in a run of open releases that also includes Fermion Research's Phonon-2 speech model and Nvidia's Kumo Tabular.
The US Federal Trade Commission is investigating Anthropic, OpenAI and other AI labs over the risks their agents pose to consumers, the first official US enforcement action on rogue AI agents, as reported on 30 September.
OpenAI and Synopsys signed a multi-year deal on 30 September to build GPT-Synopsys, a specialised model for chip design. It was one of at least five open or openly described model releases and partnerships to land in the final days of September.
OpenAI said on Tuesday it will not release its newest model, GPT-6.1 Astra, because it "didn't quite meet the bar" on safety, and it has paused training of its most powerful systems, according to CBS News and Rest of World. The same week, the UK's AI Security Institute published an open-source evaluation framework and researchers found a Chinese model would talk about bioweapons.
Fermion Research released Phonon-2 on 30 September, an open English speech recognition model that fits in a 164 MB download and averages 5.21 percent word error across the Open ASR Leaderboard's seven English sets. It landed in a week when open-weight releases, not frontier launches, did most of the talking.