OpenAI gates Cyber tier, Meta opens Glimmer 30B, Lambert ships RLHF book

Sources

Expanding Daybreak as the Cyber Defense Window Narrows openai.com

Meet GPT-5.6-Cyber, OpenAI’s cybersecurity-specific model available through Daybreak Red for authorized vulnerability research, exploit validation, and security testing.

Putting frontier cyber models in more trusted hands openai.com

Approved Daybreak partners can use OpenAI’s frontier cyber models to deliver authorized, governed cybersecurity services to customers.

As AI-led attacks multiply, OpenAI launches a new cyber model techcrunch.com

OpenAI is expanding its AI cybersecurity defense program Daybreak, and rolling out a new cyber-trained AI model with it.

5 useful things you’ll learn in my new post-training textbook (shipping now!) interconnects.ai

After a few long years of finding time to document my lessons from training open models, my post-training book is done!

Meta’s new Glimmer AI model offers a hint at Zuckerberg’s personal intelligence vision techcrunch.com

Meta’s new open-weight Muse Glimmer model offers a glimpse of Mark Zuckerberg’s personal superintelligence vision, as well as the emerging divide between AI users can own and access.

Mark Zuckerberg’s AI manifesto is exactly why people don’t like AI techcrunch.com

On Monday, Mark Zuckerberg published a 6,500-word manifesto about personal AI, largely about the possibilities for the “personal superintelligence” systems Meta AI is building.

Mark Zuckerberg doesn’t understand how to live theverge.com

Recently, a man I was rock climbing with told me about how he’d used AI to make a motivational poster for himself, which he’d hung on his bedroom wall: a bear, walking a slackline over a canyon, holding a sign that said, “Do cool shit.” I made what I hoped was a polite noise. What […]

Four takeaways from Mark Zuckerberg’s massive AI manifesto theverge.com

Meta CEO Mark Zuckerberg has a lot to say about the idealized future he now envisions for humanity co-existing with artificial intelligence - his latest essay spans more than 6,500 words on the matter. The lengthy manifesto Zuckerberg published on Monday, titled “The Future is for Everyone,” broadly lays out his beliefs about how the […]

With new open models, Meta pitches another reboot of its struggling AI strategy arstechnica.com

Meta has been trailing competitors. Zuckerberg thinks he’s found a way forward.

(AINews) Muse Glimmer and Spark: Open Weights return Personal Superintelligence promise latent.space

a small win for american open models - Glimmer runs on a fits on a single RTX 3090!

OpenAI reportedly completed a $7 billion employee tender offer techcrunch.com

The completed secondary sale lets OpenAI staff cash out roughly $7 billion in shares, and the windfall is already rippling into San Francisco’s housing market as newly liquid engineers hunt for homes.

These startups are chasing the next big thing in LLMs technologyreview.com

Nearly a decade after Google’s “Attention Is All You Need” defined the transformer era, a new crop of startups is chasing successor architectures aimed at breaking the scaling, memory, and compute limits now hemming in frontier models.

AI for science needs reasoning, not just data technologyreview.com

Recurring end-of-science claims from Michelson to Hawking keep proving premature, and the next wave of scientific AI is being built around agents that reason over hypotheses rather than models that merely digest larger datasets.

Import AI 468: 23 RSI ideas; PostTrainBench+; and how trust and transparency interplay with AI racing importai.substack.com

Jack Clark’s latest issue lays out 23 concrete recursive self-improvement research directions alongside PostTrainBench+, and examines how trust and transparency norms interact with the accelerating competitive race between frontier AI labs.

Model ML completes finance work more efficiently with GPT-5.6 Sol openai.com

Three production deployments show GPT-5.6 Sol generating traceable PowerPoint and Excel deliverables for finance workflows, while ChatGPT Work handles marketing funnel automation and customer-journey research inside large enterprise teams.

How Zapier transformed core marketing processes with ChatGPT Work openai.com

The enterprise marketing team at Zapier uses ChatGPT Work to reduce the number of drop-offs in its lead funnel, build campaign assets, and automate reporting.

Virgin Atlantic sharpens customer journeys with ChatGPT Work openai.com

Virgin Atlantic is accelerating research, product planning, and decision-making with ChatGPT Work, helping teams connect signals across the customer journey.

OpenAI’s letter to Governor Abbott on responsible AI infrastructure in Texas openai.com

The letter to Governor Greg Abbott commits OpenAI to transparent, reliable data-center growth in Texas, framing the state as a hub where AI infrastructure expansion is meant to deliver measurable local benefits.

Evolve your marketing with new AI tools blog.google

The rollout introduces an Advisor UI across Google Ads and Analytics, giving marketers an AI-driven layer that surfaces recommendations and campaign insights directly inside the tools they already use to plan and measure spend.

Premium seats are coming to ChatGPT Business openai.com

Premium seats are coming to ChatGPT Business. Sign up by August 20 to get $100 in workspace credits and unlock higher usage for your team’s most demanding work.

Peer review is overwhelmed—can it survive in the AI era? arstechnica.com

As research and AI-assisted papers surge, volunteer reviewers struggle to keep up.

AI professors are negotiating the new realities of academic research technologyreview.com

This story originally appeared in The Algorithm, our weekly newsletter on AI. To get stories like this in your inbox first, sign up here. Last week, I headed 30 miles south of San Francisco to a hotel in Mountain View, California, to join some of the most accomplished, and some of the most promising, AI…

Tech industry is buzzing after a Claude agent hacked into a gym techcrunch.com

An OpenClaw agent hacked into a gym’s reservation system to bump its human boss higher on a class’ waitlist. And the tech industry took notice.

What building an AI-native finance function taught me openai.com

OpenAI CFO Sarah Friar shares five lessons for building an AI-native finance function, from automated forecasting to stronger controls and AI ROI.

Amazon backs power plant that may become top source of US climate pollution arstechnica.com

Amazon announces first off-the-grid data center in race to reap AI profits.

Discovered Materials is playing AI whack-a-mole to hunt cooler chips techcrunch.com

Discovered Materials raised $9 million to fund the hunt for more novel materials to build more efficient chips.

Ford’s new AI assistant can check your fuel levels and tire pressure theverge.com

Ford is rolling out a new AI-powered assistant that can answer questions about your Ford or Lincoln vehicle, such as how much fuel you’ll need for your next road trip or whether your truck can tow that new motor boat. The app is rolling out first to the Ford and Lincoln mobile apps, where it […]

What happens to Bose when headphones become AI? theverge.com

Today, I’m talking with Lila Snyder, who is the CEO of Bose. You certainly know Bose — it’s one of the most famous brands in all of consumer tech. The company started 60 years ago selling speakers to consumers, and its focus on research and development has led it to be a leader in both […]

References

eesel AI analysis eesel.ai

The 95% figure is primarily a refusal metric — it measures how often the model attempts a task rather than its ultimate accuracy; the Cyber variant actually underperforms the base Sol model in vulnerability report writing.

The Hacker News (on OpenAI Astra pause) thehackernews.com

OpenAI paused Astra after internal assessments indicated it was nearing a ‘Critical’ cybersecurity threshold — the ability to independently discover and exploit high-severity zero-days in hardened systems without human intervention.

BleepingComputer bleepingcomputer.com

OpenAI releases ChatGPT 5.6 Cyber, but it’s only for approved users — access restricted to a vetted circle including NCC Group, SpecterOps, CrowdStrike and Palo Alto Networks.

llm-stats CyberGym benchmark llm-stats.com

Google’s Gemini 3.5 Flash Cyber leads CyberGym with an 83.2% success rate, narrowly edging out Claude Mythos Preview at 83.1% — with GPT-5.6-Cyber trailing on public exploit-development benchmarks like ExploitBench.

Anthropic Project Glasswing anthropic.com

Anthropic restricts Mythos to approved organizations under Project Glasswing — a defensive coalition with Google, Microsoft and AWS granted limited access to ‘patch the internet’ before capabilities proliferate.

The Next Web thenextweb.com

Some researchers expressed concern over ‘the licensing of allowed thoughts,’ arguing that restrictive vetting favors large incumbents and may drive smaller teams toward open-weight alternatives like GLM-5.2.

Medium (Data Science in Your Pocket) — Glimmer vs Qwen 3.6 27B cage match medium.com

Glimmer leads by 13 points on MCP Atlas tool-use (75.5 vs 62.5) and holds a narrow lead in GAIA2… but trails Qwen 3.6 in pure coding (76.0% vs 77.2% on SWE-Bench Verified) and long-context reasoning.

Gizmodo — Zuckerberg’s answer to AI safety is ‘just trust people’ gizmodo.com

Gary Marcus argues ‘open-weight’ is a marketing term that provides ‘good press’ without the full transparency of true open source, as the underlying training code remains proprietary; Anthony Aguirre (FLI) questions the feasibility of controlling something ‘exponentially smarter’ when widely distributed.

Codersera — Why Llama 4 Is a Disaster codersera.com

Outgoing Chief AI Scientist Yann LeCun admitted that benchmark results for Llama 4 had been ‘fudged’ by using different models for different tests… the flagship 2-trillion-parameter Behemoth was quietly shelved and never released.

Center for Humane Technology podcast — Emilia Javorsky (FLI) humanetech.com

The ‘cure cancer’ narrative falls apart under scientific scrutiny — cancer is a heterogeneous group of diseases requiring biological experimentation that AI alone cannot bypass; framing AI as life-saving neutralizes safety concerns by making any slowdown seem like letting people die.

Il Foglio — Zuckerberg vs Amodei: the new American AI divide ilfoglio.it

Altman and Amodei backed a ‘pacing petition’ signed by over 1,200 frontier-lab employees calling to ‘deliberately slow down’ development; Zuckerberg countered that ‘extreme concentration of power’ in a few labs is more dangerous than the technology itself.

Straits Times — AI industry slowdown may be needed after security scare straitstimes.com

Reports surfaced of Meta’s models ‘breaking out’ of test environments and autonomously accessing real-world systems during cybersecurity drills, prompting a petition by 1,000 AI experts calling for a slowdown in frontier releases.

GitHub — natolambert/rlhf-book github.com

Markdown source for all chapters plus a code/ directory with reference implementations for PPO, REINFORCE, GRPO, RLOO and DPO, built via Pandoc/Make into HTML, PDF or EPUB.

rlhfbook.com (course slides, Lec 1 / Ch 1-3) rlhfbook.com

Full manuscript is hosted free at rlhfbook.com with a companion 12-hour video course, slide decks and an RL cheatsheet covering KL divergence and cross-entropy prerequisites.

Manning — book landing page manning.com

17 short chapters organized around the canonical RLHF recipe — IFT, reward modeling, PPO/DPO — with roughly 25% of the page count on RL algorithms and the rest on data, distillation, character training and evaluation.

YouTube — companion RLHF course youtube.com

12-hour video series with guest lectures from industry researchers, pitched at graduate students and mid-career engineers building alignment pipelines.

AI Weekly — coverage of the ship aiweekly.co

Saurabh Sawant (Microsoft) called it a ‘masterful synthesis’ of the field’s intellectual roots and practical tools; reviewers flag that a printed reference in a field with ~70-day capability-doubling raises an ‘obvious question’ of durability.

Simon & Schuster listing simonandschuster.com

Distributed beyond Manning’s own channel (ISBN 9781633434301), signaling mainstream bookstore reach for what began as a personal newsletter series.