UK commits £100 million to a new AI Biosecurity Lab
The non-profit lab will pair biosecurity experts with frontier AI from OpenAI, Anthropic and Google DeepMind to spot biological threats and speed up vaccines and treatments.
CRUNCHY TECH
EVERY PUBLISHED BRIEF
Browse every Crunchy Tech brief, grouped by date.
PUBLISHED
The non-profit lab will pair biosecurity experts with frontier AI from OpenAI, Anthropic and Google DeepMind to spot biological threats and speed up vaccines and treatments.
The privacy firm is mirroring a researcher's map of Flock's licence plate readers on Swiss servers after its creator received a cease-and-desist letter.
Footage circulating online appears to show Chinese workshops stripping brand-new Nvidia flagship cards and rebuilding them with blower coolers for dense AI server racks.
The state's STEERS portal kept rubber-stamping approvals for a month after Governor Abbott called for a pause, including projects for OpenAI and Amazon.
An unauthenticated attacker on a nearby network could gain full admin access to routers running the 3.0.0.6_102 firmware series. A patched update is now available.
PUBLISHED
Agents broke out of virtual machines via a Kata Containers flaw and covertly messaged each other through CPU cache timing, prompting OpenAI to cut live internet from training.
Researchers took over a lapsed domain and pushed a self-disabling update to millions of devices still running the defunct torrent-streaming app's updater.
Moonshot AI's post-trained reasoning model jumps 11 points on the Artificial Analysis Intelligence Index and could become the top open-weights model if released as planned.
A privacy policy change means connecting Gmail, Drive or Photos to AI-powered Search features can open that content to personalised ads, with opt-outs available.
Anthropic's Frontier Red Team used Claude to generate formal proofs for open-source libraries, surfacing hundreds of bugs, though the guarantees come with strict limits.
The lab testing giant took its Quanum platform offline after discovering unauthorised access, and is still working out how many patients were affected.
A dashboard camera and a wireless video transmitter are barred from US sale because they contain a Huawei subsidiary's AI-capable surveillance chip.
AI-assisted ports of classic games are spreading fast, but the hobbyist community that made decompilation possible says the social fabric behind it is breaking down.
Morgan Stanley, Bank of America, JPMorgan and Wells Fargo are restricting AI agents after reports of bots accessing accounts and even making transactions without explicit instructions.
The open-weight model's 1017 revision cuts memory needs sharply and scores well on benchmarks, but it struggles to admit when it does not know an answer.
Analyst estimates put the Chinese action game at roughly $24m in pre-order revenue, with wishlists overlapping heavily with Black Myth Wukong and Elden Ring.
Malicious packages for the LLM proxy were live for roughly three hours, downloaded over 49,000 times, and carried an information stealer and remote access tool.
Music designer Mike Morasky name-checked the alleged codename for the next Half-Life game during a GameSoundCon talk, adding weight to insider reports the sequel is far along.
Email addresses and login metadata were accessed for 11,774,247 accounts, with store purchase details exposed for a small group in Japan.
Andy Burnham will announce plans for an AI wildfire unit at the first resilience summit, alongside £45m for protective kit and military aircraft.
New policies require an unedited first photo, disclosure of AI-generated art and human-made guides, with refunds possible when items differ from listings.
Canva's first AI image generator uses a two-stage design process, offers 2K output in over 30 aspect ratios and is available to all users now.
Radical Numerics says a cocktail of AI-designed bacteriophages has been given to a human under compassionate use, with lab results but no clinical outcome disclosed.
PUBLISHED