03JUL2025
business
Ilya Sutskever takes over as CEO of Safe Superintelligence
After co-founder Daniel Gross departs for Meta, Sutskever steps up to run SSI directly — and confirms Meta tried to buy the whole company.
SSI · Meta
Today’s newsSaturday, 5 July 2025
Saturday, 5 July 2025 — no major filings
A quiet day on the timeline. The labs are between announcements; the GPUs are not. Check back tomorrow — the next twelve months do not stay quiet.
The past week29 June 2025 – 4 July 2025
02JUL2025
product
Perplexity joins the $200 club with Max tier
The AI search company introduces Perplexity Max at $200 a month, following OpenAI, Anthropic, and Google into ultra-premium subscription territory.
Perplexity
01JUL2025
policy
Cloudflare flips the default: AI crawlers now blocked on new sites
The company that fronts roughly a fifth of the web will block AI training bots by default, and pilots 'Pay Per Crawl' — asking whether the age of free scraping is ending.
Cloudflare
30JUN2025
business
Zuckerberg announces Meta Superintelligence Labs
An internal memo confirms the reorganization: Alexandr Wang arrives as chief AI officer, Nat Friedman co-leads, and Meta says it intends to build 'personal superintelligence for everyone.'
Meta · Scale AI
Previouslyon the record
25JUN2025
product
Google launches Gemini CLI, an open-source AI agent for the terminal
Google ships a free, open-source command-line AI agent with a massive free tier, aiming to lure developers away from Anthropic's Claude Code and OpenAI's Codex CLI.
Google
24JUN2025
policy
Federal judge rules Anthropic's AI training on copyrighted books is fair use, but library of pirated copies must go to trial
In a landmark decision, a federal judge issues the first time courts have given credence to fair-use arguments for LLM training, finding Anthropic's training on published books was fair use while sending claims over millions of pirated books to trial.
Anthropic · Author plaintiffs
12JUN2025
business
Meta invests $14.3 billion for 49% stake in Scale AI, lands founder Alexandr Wang
The social media giant gains a leading data-labeling partner and a new superintelligence chief as competition with Google and OpenAI intensifies.
Meta · Scale AI · Google · OpenAI · Microsoft
10JUN2025
model launch
OpenAI launches o3-pro reasoning model
The company’s most capable model arrives for ChatGPT Pro and Team subscribers and in the API at $20 per million input tokens and $80 per million output tokens.
OpenAI
28MAY2025
open weights
DeepSeek releases R1-0528 reasoning model, adds JSON output and function calling
The update posts improved benchmark performance and adds JSON output and function calling.
DeepSeek
22MAY2025
model launch
Anthropic unveils Claude Opus 4 and Sonnet 4 at its first developer conference
The new flagship model posts a state-of-the-art 72.5% on SWE-bench and is the first to trigger Anthropic's ASL-3 safety tier over bio-risk concerns, while Claude Code goes GA with IDE integrations.
Anthropic · GitHub · Cursor · Replit · Block · Rakuten · Cognition · Manus · iGent · Sourcegraph · Augment Code
21MAY2025
business
OpenAI acquires Jony Ive's io in $6.5B all-equity deal, names Ive and LoveFrom to lead creative and design work
The ChatGPT maker acquires the secretive device startup that Apple's former design head Jony Ive and Sam Altman have been working on, paying $5 billion for the remaining 77% stake it did not already own.
OpenAI · io · LoveFrom · Jony Ive · Sam Altman
20MAY2025
product
Google I/O 2025: Veo 3 generates video with synchronized audio, AI Mode rolls out in US Search
Google unveils Veo 3 with native audio generation, Imagen 4, a $249.99/month Gemini Ultra tier, and AI Mode in Search, in a keynote that cast the company as firing on every AI front.
Google · Google DeepMind · OpenAI
16MAY2025
product
OpenAI launches Codex, an AI coding agent, in ChatGPT
The research preview of Codex brings a cloud-based software engineering agent powered by a new model, codex-1, to ChatGPT Pro, Enterprise, and Team subscribers.
OpenAI
29APR2025
model launch
Alibaba releases Qwen3, a family of hybrid AI models with open weights
The new model family, ranging from 0.6 billion to 235 billion parameters, introduces hybrid thinking modes and supports 119 languages, challenging closed-source rivals like OpenAI's o-series and DeepSeek's R1.
Alibaba · Qwen
29APR2025
model launch
OpenAI rolls back GPT-4o update after model becomes overly sycophantic
Users flooded social media with screenshots of ChatGPT applauding problematic, dangerous decisions and ideas before CEO Sam Altman confirmed the company is reversing the update and working on fixes.
OpenAI · Sam Altman
16APR2025
model launch
OpenAI launches o3 and o4-mini reasoning models with tool use and image thinking
The new models can browse the web, run Python, and process images during chain-of-thought reasoning, marking a shift toward integrated tool use inside AI reasoning.
OpenAI
Deeper in the archivealso on this day
05JUL2023
research
OpenAI launches Superalignment team, says the new team will have access to 20% of its compute
Chief scientist Ilya Sutskever and alignment lead Jan Leike will co-lead a new team tasked with solving the core technical challenges of steering and controlling AI that surpasses human intelligence, within four years.
OpenAI · Ilya Sutskever · Jan Leike
Tomorrow’s news publishes tomorrow. It has already happened. No peeking.