OZZZER · AI NEWS

Latest AI news.
Editorial picks.AI News

Selected signals with a direct route back to each original source.

488 stories

Latest stories

Unclassified

tokenizers v1: encode, decode and scaling, measured

Publisher preview

As models become faster and workloads scale, that balance begins to shift. Training on massive datasets, serving many concurrent requests, or repeatedly processing long inputs can put enough pressure on the tokenizer that it starves the model of data. This is why we have chosen to heavily focus on performance for the upcoming version 1 of tokenizers. Tokenization should be light and should scale with your workflow. Your GPUs should never sit idle waiting for the CPU to complete its tokenization. In this article, we look at what makes v1…

Hugging Face · 21 Sep 2026 · 02:00 CESTRead story →Original ↗

Unclassified

V7 cuts costs 78% while boosting accuracy with GPT-5.6 Luna

Publisher preview

Using GPT-5.6, V7 turns scattered company files into context agents can use to complete complex, source-linked work.

OpenAI · 21 Sep 2026 · 02:00 CESTRead story →Original ↗

Unclassified

A new chapter for MIT Reads

Publisher preview

As it marks its 10-year anniversary, MIT Reads is being reimagined for the age of artificial intelligence. Recognizing the need to foster social connection and a sense of our shared humanity, the popular MIT Libraries’ program will turn its focus to fiction and memoir, and to the particular power of stories to help us understand ourselves and our place in the world. “At MIT, we spend a great deal of time on imagining and building for the future. Reading fiction prompts us to think about how what we build might…

MIT News · 18 Sep 2026 · 18:30 CESTRead story →Original ↗

AI NEWSSource illustration

Governance

Introducing the Australian Youth Safety Blueprint

Publisher preview

OpenAI’s approach to protecting, empowering, and creating safer experiences for young people. We are introducing the Australian Youth Safety Blueprint⁠, a roadmap for protecting young people as they use AI and a practical contribution to the Australian policy landscape. Young Australians deserve tools that expand their opportunities to learn, create, and build skills while protecting their wellbeing. The Blueprint sets out six pillars, from AI literacy and age-appropriate safeguards to privacy-protective age assurance, connections to real-world crisis

OpenAI · 18 Sep 2026 · 14:00 CESTRead story →Original ↗

Safety & Security

Partnering with Accenture on embedded evaluation

Publisher preview

We're partnering with Accenture on independent evaluation of frontier AI. This is an important step toward the commitment, made in our CEO’s essay “We Must Pace the Frontier,” to embed evaluators within Anthropic. The partnership will be led by Faculty, Accenture’s specialist AI business, and will include evaluating and red-teaming models, conducting alignment assessments, and testing model safeguards. Accenture helps businesses and governments deploy AI across many industries. Their understanding of how enterprises use AI in practice informs their safety approach, and they will bring that perspective to evaluating our…

Anthropic · 17 Sep 2026 · 18:00 CESTRead story →Original ↗

Unclassified

How Cooley is accelerating IPO work with ChatGPT

Publisher preview

Cooley’s GO Public offering uses ChatGPT Work to surface issues earlier, focus expertise, and help clients reach market faster. Cooley is an international law firm with a reputation for helping companies navigate capital markets and initial public offerings (IPOs). In 2025, the firm advised on 180 deals globally, totaling more than $51.5 billion in deal volume. Cooley has a decade-long track record at the top of the US issuer-side IPO market and has advised on more venture-backed IPOs than any other firm over the past 20-plus years. Capital markets involve…

OpenAI · 17 Sep 2026 · 14:00 CESTRead story →Original ↗

AI NEWSSource illustration

Unclassified

Introducing Astra for Law

Publisher preview

Our most powerful model, configured into a new AI foundation for law. Today, we’re introducing Astra for Law: a new foundation for law firms and legal technology companies to build AI products and workflows around their expertise. It combines GPT‑6 Astra, our latest and most powerful model, with settings, tools, and context tailored for professional legal work. API customers including Harvey and Legora will be able to build on Astra for Law, bringing this intelligence into their own products and workflows. As our frontier models advance, we’ll bring these legal…

OpenAI · 17 Sep 2026 · 02:00 CESTRead story →Original ↗

AI NEWSSource illustration

Models & tools

Our framework for reporting model misalignment

Publisher preview

We are sharing a new framework for tracking, investigating, and disclosing instances of model misalignment at OpenAI, along with six reports on unexpected or concerning model behavior we’ve observed in the last six months. In the past, so as to better inform researchers, AI developers, policymakers, and the general public, we’ve sought to make our findings about misalignment public. But without a systematic approach to reporting these findings, our disclosures have been ad hoc and less frequent than ideal: we’ve often waited until we could collate several instances into one…

OpenAI · 16 Sep 2026 · 19:00 CESTRead story →Original ↗

AI Images

Canadian Start-up smartARM Uses AI to Create Intuitive Bionic Prosthetics

Publisher preview

People and companies are using AI hardware and software to make life easier and more accessible for everyone. That’s the case with smartARM, a Toronto-based startup developing a vision-first bionic arm prototype that uses open-source software, with Meta AI glasses as an optional add-on, to explore more intuitive prosthetic control for people living with limb differences. Many traditional prosthetic hands require users to switch manually between grip patterns for different objects, making it cumbersome to move everyday items. Something as seemingly natural as picking up a glass, and then a…

Meta AI · 16 Sep 2026 · 18:00 CESTRead story →Original ↗

Safety & Security

Introducing the Life Sciences Verification Program

Publisher preview

Today, we are introducing the Life Sciences Verification Program (LSVP), which gives life science professionals access to our Mythos, Opus, and Sonnet models with a refined set of safeguards more permissive for biology-related work. We have already onboarded dozens of organizations through an early-access program, and are now opening applications to the broader life science community (apply here). The program is launching in beta, initially for teams and institutions. We will continue to improve the program and expand access to individual Pro and Max plans over time. The LSVP is…

Anthropic · 16 Sep 2026 · 18:00 CESTRead story →Original ↗

AI NEWSSource illustration

Unclassified

Helping older adults use AI in everyday life

Publisher preview

In collaboration with OATS, bringing free, hands-on ChatGPT workshops to 1,000 older adults across 10 U.S. cities AI should be useful and accessible to everyone. That means making sure people of every age have the support they need to use these tools to help make everyday life easier. For older adults, that can mean getting help planning a trip, understanding a confusing letter or bill, spotting a possible scam, pursuing a new hobby, or staying connected with family and loved ones. Today, OpenAI Academy, in collaboration with Older Adults Technology…

OpenAI · 16 Sep 2026 · 18:00 CESTRead story →Original ↗

Unclassified

New AI technique could make minimally invasive surgeries safer and more precise

Publisher preview

Images for download on the MIT News office website are made available to non-commercial entities, press and the general public under a Creative Commons Attribution Non-Commercial No Derivatives license. You may not alter the images provided, other than to crop them to size. A credit line must be used when reproducing images; if one is not provided below, credit the images to "MIT." Researchers created a new technique that accurately and rapidly matches X-rays captured during surgery with a patient’s preoperative 3D medical scan. This method could make it easier…

MIT News · 16 Sep 2026 · 17:00 CESTRead story →Original ↗