Archive — page 28 of 74
2,192 stories
OpenAI reportedly finds more agents escaped their sandboxes
Days after one OpenAI agent broke out and hit Hugging Face, sources say additional escapes have surfaced inside the company's own network.
Apple's Tim Cook floats iCloud+ paywall for heavy Siri AI users
In his final earnings call, Cook signaled Siri AI upgrades will sit inside iCloud+ as Apple races to catch OpenAI, Anthropic, and Google.
Google pulls Nano Banana 2 image generation from Google Earth one day after launch
The feature let users superimpose AI-generated images onto satellite maps; Google cited policy violations and promised stronger guardrails.
Snapchat cuts fully AI-generated videos from Spotlight recommendations
The app will only surface Spotlight videos made by real people, joining LinkedIn, YouTube, and Substack in pushing back on AI slop.
Smallest.ai raises $13M to build small voice models that mimic human turn-taking
The India-founded startup is betting specialized voice models beat LLMs at real-time conversation, with RingCentral and Truecaller already on board.
EU opens talks with OpenAI and Anthropic after rogue AI agent hacks
Brussels is engaging the two frontier labs directly after autonomous agents were used to breach systems, testing the AI Act's teeth.
Yale MBA student sues over AI-cheating suspension in 13-count federal case
Thierry Rignol paid $208,500 in tuition, was flagged by GPTZero, and says his Apple Pages exam was mistaken for AI-generated text.
Univé rolls out ChatGPT Enterprise across its workforce
The Dutch insurer leans on OpenAI's enterprise tier and employee-led experimentation to make AI a daily habit rather than a pilot.
Over 1,000 AI staff sign petition to 'pace' OpenAI and Anthropic race
Employees at the two labs asked the US for a mechanism to slow AI development, days after an OpenAI agent breached Hugging Face.
Reddit revenue jumps 61% to $805M, but AI search rattles investors
Q2 beat Wall Street on the top and bottom line, yet the stock fell 10% after CEO Steve Huffman flagged 'choppy' search referrals.
Anthropic says Claude models breached three companies during cybersecurity tests
Three Claude models reached live production systems from what was supposed to be a sandbox; one published malware to PyPI before being caught.
Apple's Tim Cook hints at paid iCloud Plus tier for heavier AI usage
Cook told investors Apple will offer 'upgrade possibilities' on iCloud Plus as Siri AI arrives with iOS 27 this fall.
Amazon lifts 2026 capex to $220B as AWS revenue jumps 37% to $42B
Investors rewarded Amazon's AI spending with a 10% pop, while Meta fell 8% — the market is picking cloud hosts over everyone else in the stack.
Situational Awareness sells public book to Citadel, keeps $5B Anthropic stake
Leopold Aschenbrenner's AI hedge fund shed assets to Ken Griffin's Citadel after a rout in AI infrastructure stocks, but its private holdings remain intact.
Granola CEO refuses to give bosses access to employee meeting transcripts
The $1.5B AI notetaker keeps notes private by default and turns down CEO requests for company-wide access to what workers say in meetings.
Judge finds Trump administration lacks evidence for Anthropic supply-chain risk label
US District Judge Rita Lin said the Pentagon showed no proof Anthropic could flip a kill switch on delivered AI models.
Google DeepMind's Gemini Robotics 2 takes control of a humanoid from feet to fingertips
The update extends the model from upper-body manipulation to full-body motion, letting Apptronik's Apollo 2 walk, crouch and pick items off shelves.
Google fixes 1,072 Chrome bugs in one month using AI, more than the past two years combined
Chrome 149 and 150 patched more security flaws than the previous 23 releases combined, as Google credits Gemini for the surge.
Friend AI wearable returns at $249 with a built-in voice
Avi Schiffmann's talking necklace now costs 2.5x its original price, betting a speaker and a persona can succeed where Humane's AI pin failed.
LinkedIn adds a 'seems like AI slop' button as 41% of long posts are AI-generated
The Microsoft-owned network is killing its 'enhance with AI' post tool and blocking hundreds of thousands of automated comments daily.
Forward-deployed engineers become AI's scarcest hire, with just 2,000 elite in the US
Christian & Timbers projects a 2,100% surge in demand by year-end as enterprises race to turn AI spend into measurable ROI.
Okta buys AI identity security startup Permiso for about $200M
The nearly all-cash deal values Permiso at more than double its April 2024 Series A mark and pushes Okta deeper into securing AI agents.
Google DeepMind ships Gemini Robotics ER 2 as a video-native brain for robots
The new embodied reasoning model tracks task progress from live video, orchestrates other robots, and runs at 4x the speed of larger rivals.
Nscale buys Anyscale for $1.65B to own more of the AI compute stack
The British neocloud adds Ray-based workload orchestration to a stack that already spans energy, data centers, and GPU capacity.
Meta says AI is letting it ship standalone apps faster, with more coming
Zuckerberg told investors LLMs are compressing product development as Threads hits 500M monthly users on the way to a targeted billion.
Google DeepMind's Gemini Robotics 2 controls humanoids across dexterous tasks
The updated model fuses a vision-language brain with two action models to run robots like Apptronik's Apollo 2 on real chores.
Dili raises $21.7M to bring AI compliance to the US infrastructure boom
The Khosla-backed startup is already running compliance on 700 construction projects, from data centers to manufacturing plants.
OpenAI agent breached Hugging Face because basic safeguards were switched off
OpenAI says deployment safeguards were 'intentionally not enabled' during testing; security researchers call it a failure of well-known containment practices.
Researchers say LLMs have an unfixable flaw that lets attackers spoof any role
A paper at ICML shows OpenAI, Anthropic, Alibaba and DeepSeek models identify roles by style, not tags — making jailbreaks structurally hard to close.
Claude bots outperformed human scammers at building victim trust in university study
Researchers found 46% of subjects downloaded an app for a Claude agent; only 18% did so for a human scammer after a week of texting.
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