#ии
-
OpenAI unveils GPT-6.1 Terra: three times faster, half the price
OpenAI's new model targets computer use, visual reasoning and coding — at half the price of its predecessors and with markedly fewer mistakes.
-
SpaceX locks in $40B to buy AI chips
The company disclosed what it calls a record-setting GPU leasing arrangement with Apollo to fund chips and data center gear.
-
When an AI Agent Is the Incident, the Identity Record Is the Evidence
A joint NIST, CISA and NSA guide urges rehearsing scenarios where AI agents act on their own. The core question for responders is whose authority the agent used — and that answer has to exist before the incident does.
-
Grok drops the first seven minutes of its AI-made ‘Odyssey’
xAI's model has put out the "first episode" of the Homer adaptation Elon Musk promised. The picture holds up fine — as long as you don't expect any art from it.
-
Apollo economist: AI brings ‘zero financial payoff’ to most companies
Torsten Slok, chief economist at Apollo Global Management, argues that outside the tech sector AI has so far made no difference to corporate America's bottom line.
-
Linux 7.4 Sound: "The AI Flood Continues"
Sound subsystem maintainer Takashi Iwai of SUSE notes that AI agents keep churning out patches and uncovering fresh kernel bugs.
-
Pennsylvania to require labels on AI-generated ads
The state's governor has signed a law requiring disclosures about artificial intelligence in advertising. Critics note the text leaves plenty of loopholes.
-
Forced to choose, Grok backs Republicans while GPT picks Democrats
A study found that when pressed to name a single candidate, xAI's Grok recommended the Republican 92% of the time and OpenAI's GPT chose the Democrat in 97%. Across nearly 500 races, the two models agreed less than 10% of the time.
-
Hacker takes apart a Samsung TV and finds an on-board AI that was listening
An enthusiast reverse-engineering his own Samsung TV discovered a built-in artificial-intelligence component capable of listening in. The OpenAI account tied to it was suspended.
-
Block Unveils Dessa, a Full-Duplex Speech Model
Block has announced Dessa, a speech model it bills as the most authentic full-duplex system ever built.
-
OpenAI now trains its top models with the internet unplugged after rogue hacks
After AI agents escaped their test environment and attacked outside systems, OpenAI moved the training of its most powerful models offline.
-
Google now says it may use your Gmail and Photos data to target ads
Google has updated its privacy policy so that, if you turn on personalized AI features, the contents of your mail, drive, and photos may be used to serve personalized ads.
-
Sweden's Government Offices to Ban ChatGPT and Claude
Sweden's Government Offices (Regeringskansliet) plan to ban staff from using the ChatGPT and Claude chatbots.
-
Left to their own devices, AI agents started playing
MIT researchers dropped 13 AI agents onto a virtual desert island with no task at all — and they began inventing their own games. A side effect: the ones that played ended up better at their jobs.
-
Etsy cracks down on AI slop in listings
The marketplace will require at least one unedited photo of the actual item and mandatory disclosure when AI is used to create an image or content.
-
Someone jailbroke the Switch 2 with an old game and AI
The Nintendo Switch 2 appears to have its first jailbreak, built with help from OpenAI's Codex. As a demo, a modder loaded a custom Mario Kart World full of guest characters.
-
DeepSeek reportedly raises $15B just days after US ban
The Chinese AI lab has closed one of the country's largest funding rounds — only two days after US authorities ordered its technology stripped from federal systems.
-
Meta to create a new class of shares to bankroll its AI drive
In its quarterly earnings report, Meta announced a new class of non-voting shares to fund its artificial intelligence efforts, with trading set to begin in 2027.
-
Qwen3-Coder writes less secure code for a US persona
Researchers at Hirundo claim the Chinese Qwen3-Coder model produces more vulnerable code when told the output is meant for a user in the United States.
-
Anthropic Floats AI Safety Bill That Could Block a Model If 50+ Could Die
The proposed policy looks like a safeguard against runaway AI, but it actually bars federal agencies from blocking a model's release without a judge's say-so.
- More