AI Today: OpenAI DevDay 2026, GPT-6.1 Sol, SynthID Bio, FTC Probe & More – 30 September 2026

Here is everything that mattered in AI today, Wednesday 30 September 2026.

OpenAI DevDay 2026: New Models, APIs, and 1.2 Billion Weekly Active Users

OpenAI held its most confident DevDay yet, unveiling a raft of new products and milestones. Highlights from the event include the introduction of Dots, the GPT-6.1 Sol model, an Ultrafast inference tier, a Decisions API, an Agents API, Spaces, and a developer Marketplace. Perhaps the most striking statistic shared was that ChatGPT has now reached 1.2 billion weekly active users — a figure that underscores just how central the platform has become to everyday AI use. The breadth of announcements signals OpenAI's push to become not just a model provider but a full-stack platform for agentic and consumer AI applications.

GPT-6.1 Sol Replaces GPT-6 Sol After Just Seven Days

In a sign of how rapidly the frontier is moving, GPT-6.1 Sol has already replaced GPT-6 Sol — only seven days after the latter's release. According to analysis from Artificial Analysis, the new model delivers near-Astra-level intelligence, representing a meaningful capability jump in an extremely short turnaround. The rapid iteration reflects the intense competitive pressure across the industry and raises questions about how developers and enterprises can keep pace with model versioning at this speed.

OpenAI Disrupts Coordinated Model-Distillation Campaign

OpenAI published details of its efforts to disrupt a coordinated campaign aimed at extracting protected model reasoning through adversarial distillation. The company explained how bad actors attempted to systematically draw out proprietary reasoning patterns — effectively stealing model capabilities — and outlined the defensive measures now being strengthened to prevent such attacks in future. The disclosure highlights a growing category of threat that sits at the intersection of IP protection and AI security, and is likely to prompt broader industry discussion about safeguarding frontier model outputs.

Google DeepMind Introduces SynthID Bio

Google DeepMind announced SynthID Bio, a proof-of-concept system for watermarking AI-generated proteins while preserving their biological function. Building on the SynthID watermarking framework previously applied to text and images, the new tool extends the concept into the life sciences — embedding verifiable markers into synthetic protein sequences without degrading their utility. The announcement is significant for biosecurity and scientific integrity, as AI-generated biological designs become increasingly sophisticated and widespread.

FTC Opens Probe into OpenAI, Anthropic, and Other AI Giants

The US Federal Trade Commission has opened an investigation into major AI companies including OpenAI and Anthropic, according to reporting cited on Hacker News. While details of the scope remain limited in the available reporting, the probe adds to a growing pattern of regulatory scrutiny directed at the AI industry's largest players. For Australian businesses and developers operating in or alongside US AI ecosystems, regulatory developments of this kind have the potential to shape the terms on which these platforms operate globally.

OpenAI Partners with America's SBDC to Support Small Businesses

On a more constructive note, OpenAI announced a partnership with America's Small Business Development Center (SBDC) network to expand hands-on AI training and local support for small businesses. The initiative is accompanied by a new report examining how small teams are already putting AI to work. The move reflects a broader effort by OpenAI to deepen adoption beyond enterprise and developer audiences and into the everyday economy — a trend relevant to small business owners everywhere, including in Australia.