Archive — page 37 of 62
1,853 stories
Probably raises $9M from a16z to push LLM accuracy to 99.99%
Founder Peter Elias says tighter harness engineering lets Probably run on models four classes weaker than the frontier, cutting token costs.
Microsoft patches critical Copilot flaw that leaked 2FA codes from user emails
Varonis's SearchLeak exploit chain weaponized a crafted URL to make M365 Copilot exfiltrate emails, SharePoint docs, and OneDrive files.
SpaceX to acquire Cursor for $60B in stock, days after record IPO
The all-stock deal closes a path Musk laid out in April, when SpaceX agreed to buy Cursor or pay a $10B break-up fee.
Tokenomics arrives: how 8x8 and Baseball Lifestyle 101 are managing Claude bills
Royal Bank of Canada's token usage jumped 500% in six months as CFOs grapple with a new line item that didn't exist two years ago.
Emerald AI's Conductor pitches flexible data centers as the grid bottleneck answer
A Duke study says US grids can absorb 76 GW more — about 5% of total capacity — if data centers flex just 0.25% of the year.
Respond.io raises $62.5M Series B as messaging AI hits $35M ARR
The Malaysia-based conversational platform grew 169% year-over-year and is hunting acquisitions in North America and Europe.
DOJ backs xAI in NAACP suit, calling Grok 'vital' to military operations
Justice Department says shutting xAI's unpermitted gas turbines at Colossus 2 would threaten Iran strike operations and national security.
Meta CTO calls company's Applied AI reorg 'atrocious' in internal memo
Andrew Bosworth pledged stability, manager caps, and better snacks after a 6,500-person AI division was labeled 'a gulag' by staff.
Nvidia raises $25B in first bond sale since 2021 as AI debt market swells
Orders topped $85B for the seven-part offering, letting Nvidia tighten pricing and triple its outstanding debt to about $30B.
Anthropic shutdown of Fable 5 and Mythos 5 fuels sovereign AI push abroad
After the White House forced Anthropic to cut foreign access to its top models on June 13, the UK, France, and Canada are using it to argue for homegrown AI.
Meta launches AI Mode on Facebook, pulling answers from public posts and Groups
Facebook's new AI Mode summarizes public posts, Groups, and Reels in response to plain-language questions — and joins a growing AI subscription push.
South Korea leads the world on AI optimism as government bets on chips and chatbots
Only 16% of South Koreans are more worried than excited about AI — the lowest of 25 countries Pew surveyed, versus 50% in the US.
White House ties AI preemption to KOSA in last-ditch federal push
With midterms five months away and a five-week recess looming, the administration is binding AI rules to child-safety law.
76 cybersecurity experts protest US export ban on Anthropic's Mythos and Fable
Alex Stamos, Katie Moussouris, and Jon Callas signed an open letter calling the White House order 'dangerous' for defenders.
Salesforce buys AI customer service platform Fin for $3.6B
The deal feeds Fin's Apex model and Operator agent into Agentforce as Salesforce shares slide more than a third in 2026.
Loft Orbital's Yam-9 runs first vision-language model in orbit
A Google DeepMind Gemma 3 model running on a Nvidia Jetson chip aboard Yam-9 identified targets from natural language queries.
NewCore raises $66M to give AI agents enterprise identities
The Israeli-American startup launches with a $300M valuation, betting legacy identity platforms can't handle a workforce of humans and software agents.
Sarvam hits $1.5B valuation with $234M round led by HCLTech
India's newest AI unicorn lands a strategic check from HCLTech as sovereign-AI pressure intensifies across markets.
AT&S commits up to €2B to Malaysia plant to ride AI chip demand
The Austrian substrate maker bets on Southeast Asia capacity to serve AI accelerator customers chasing Nvidia-class packaging.
Sequent launches to fix alignment before superintelligence arrives
A new nonprofit from UK AISI and Timaeus researchers wants $100–150M to chase theoretical alignment guarantees the frontier labs aren't pursuing.
Meta licensed Pentagon supplier Rank One's face recognition for smart glasses
A leaked software license shows Meta tested face recognition from a vendor that draws 80% of its revenue from government clients.
Tech layoffs hit 150,000 in 2026 as AI cited in record cuts
Companies are cutting workers at 974 per day while a handful of AI insiders mint billion-dollar fortunes — and the gap is widening.
White House suspects China-linked group accessed Anthropic's Mythos
Export restrictions on Anthropic's most powerful model were partly driven by suspected unauthorized access, according to a Semafor report.
Anthropic's Fable 5 suspension reignites India's sovereign AI debate
India is OpenAI and Anthropic's second-largest market — and a U.S. directive just cut off access to the newest models for foreign nationals.
Meta unwinds $2B Manus acquisition after Beijing divestiture order
Meta has cut Manus off from internal systems as the Chinese-founded startup explores a $1B buyback and Hong Kong listing.
KPMG pulls AI report after hallucinations name UBS, NHS as users
An October 2025 KPMG report on agentic AI cited UBS, the NHS, Swiss rail, and TfL as customers. They say it isn't true.
Amazon security paper triggered White House ban on Anthropic's Fable
Andy Jassy briefed officials on a cyberattack-prompt vulnerability days before the export control order that pulled Fable 5 and Mythos 5.
Amazon CEO Andy Jassy flagged Anthropic model risks to Treasury before export ban
Jassy told Scott Bessent that Amazon researchers used Claude Fable 5 to extract cyberattack-relevant info, preceding the Fable 5 and Mythos 5 shutdown.
OpenAI subpoenaed by New York AG in multistate investigation
The probe spans advertising, model sycophancy, consumer data, and treatment of minors — and lands days after OpenAI filed confidentially to IPO.
Tribeca 2026 shows AI filmmaking needs custom models, not vanilla prompts
Google DeepMind and OpenAI brought AI shorts to Tribeca. The standout used custom Veo and Imagen builds trained on hand-painted Pixar concept art.
Stay ahead of everyone in AI.
The tightly edited AI news email engineers, founders, and investors actually open. One email. Every weekday. Five minutes to finish.
The briefing read inside teams at