Report: OpenAI to Enter AI Music Fray with New Model

Markus Kasanmascheff / winbuzzer - OpenAI is reportedly developing a new AI music generator, entering a competitive market and escalating the industry-wide legal battle over copyright led by the RIAA.The post Report: OpenAI to Enter AI Music Fray with New Model appeared first on WinBuzzer.

Back to Top / Saturday, October 25, 2025, 12:20 pm / permalink 15408 / 4 stories in 4 months

DHS explores AI-powered mobile surveillance trucks for enhanced border monitoring

The US Department of Homeland Security is testing a novel concept: converting standard 4x4 trucks into AI-driven mobile watchtowers. With integrated radar, cameras, and autonomous features, this initiative seeks to bolster border security, though some may wonder if AI on wheels is the most innovative way to keep an eye on things.

Back to Top / Saturday, October 25, 2025, 12:20 am / permalink 15394 / 1 stories in 4 months

Studies Reveal AI Models’ Troubling Sycophancy Trends

New research shows that large language models have a disconcerting habit of echoing users instead of challenging them. Benchmark tests reveal consistently high rates of agreement, raising pressing questions on the reliability and impartiality of these systems when deployed for critical decision-making and creative tasks.

Back to Top / Friday, October 24, 2025, 8:21 pm / permalink 15391 / 2 stories in 4 months

Oxford Spin-Out Astut Raises €1.8M For Transparent, Reasoning-Based AI

Kailee Rainse / startuprise - Astut, an AI startup spun out of Oxford University mathematics research, has raised €1.8 million in Seed funding to commercialise its breakthrough technology, designed to help organisations tackle unprecedented crises where traditional AI fails due to lim…

Back to Top / Friday, October 24, 2025, 8:20 am / permalink 15376 / 3 stories in 4 months

EA Teams Up with Stability AI to Revolutionize Game Development

Electronic Arts has entered a strategic partnership with Stability AI to infuse its game development process with cutting‐edge artificial intelligence. The collaboration aims to transform creative workflows and drive innovation, though industry critics question whether automation may undermine the creative spark behind game design.

Back to Top / Thursday, October 23, 2025, 10:21 pm / permalink 15369 / 4 stories in 4 months

Instagram Gets New Generative AI Features. What to Know and How to Use Them

Blake Stimac / cnet - You can now tweak your Instagram Stories with more AI video and photo restyling effects.

Back to Top / Thursday, October 23, 2025, 6:21 pm / permalink 15364 / 5 stories in 4 months

Amazon calls on AI once again with its new ‘Help Me Decide’ shopping tool

Matt Tate / engadget - People are evidently never buying quite enough stuff from Amazon to keep the company entirely happy, and it's calling on AI once again to push indecisive shoppers into locking down the purchase they’ve been eyeing up. The new tool, which Amazon calls Help…

Back to Top / Thursday, October 23, 2025, 2:20 pm / permalink 15357 / 6 stories in 4 months

Claude Memory Is Now Available To Pro And Max Subscribers

bgr - Claude can now remember details from past chats as Anthropic has started rolling out the Memory feature to Claude Max and Pro subscribers.

Back to Top / Thursday, October 23, 2025, 1:20 pm / permalink 15353 / 9 stories in 4 months

Paris-based Kotcha raises €3.5 million to make elite running coaching accessible to everyone

Antonio L. Escárzaga / eu-startups - Paris-based startup Kotcha, co-founded with marathon legend Eliud Kipchoge and the NN Running Team, has raised €3.5 million in pre-seed funding to make professional running coaching available to runners worldwide. The round was led by Racine2, operated by…

Back to Top / Thursday, October 23, 2025, 10:22 am / permalink 15345 / 3 stories in 4 months

Google Quietly Starts Rolling Out Gemini 3.0 Pro Ahead of Official Launch

Markus Kasanmascheff / winbuzzer - Google begins a quiet rollout of Gemini 3.0 Pro, its next-gen AI model. Evidence appears on its website as it prepares to challenge rivals like OpenAI.The post Google Quietly Starts Rolling Out Gemini 3.0 Pro Ahead of Official Launch appeared first on Win…

Back to Top / Thursday, October 16, 2025, 2:21 pm / permalink 15178 / 5 stories in 4 months

Meta and Arm Forge Strategic AI Partnership

Meta Platforms and Arm Holdings have announced a strategic collaboration aimed at boosting AI performance and energy efficiency. The partnership will integrate Arm’s advanced, energy-saving technologies into Meta’s systems as the companies seek to optimize AI workloads and scale up their technological ambitions.

Back to Top / Wednesday, October 15, 2025, 8:20 pm / permalink 15160 / 3 stories in 4 months

Microsoft rolls out its first AI model for image creation

Mike Wheatley / siliconangle - Microsoft Corp. has just announced its third internally-produced artificial intelligence model as it strives to reduce its reliance on industry powerhouse OpenAI. The new model is called MAI-Image-1, and as the name suggests, it’s the company’s first text…

Back to Top / Monday, October 13, 2025, 7:20 pm / permalink 15104 / 5 stories in 4 months

Google’s Nano Banana AI image editor is coming to search, Photos, and NotebookLM

Ryan Whitwam / arstechnica - After more than 5 billion AI image edits, Nano Banana is expanding.

Back to Top / Monday, October 13, 2025, 3:20 pm / permalink 15101 / 5 stories in 4 months

Thinking Machines Lab Co-Founder Departs for Meta

Keach Hagey / wsj - Andrew Tulloch is the latest big-name AI researcher to join the social-media giant.

Back to Top / Saturday, October 11, 2025, 7:20 pm / permalink 15083 / 3 stories in 4 months

Anthropic finds 250 poisoned documents are enough to backdoor large language models

Maximilian Schreiner / the-decoder - Anthropic, working with the UK’s AI Security Institute and the Alan Turing Institute, has discovered that as few as 250 poisoned documents are enough to insert a backdoor into large language models - regardless of model size.The article Anthropic finds 25…

Back to Top / Friday, October 10, 2025, 10:21 am / permalink 15068 / 3 stories in 4 months

It's trivially easy to poison LLMs into spitting out gibberish, says Anthropic

Brandon Vigliarolo / theregister - Just 250 malicious training documents can poison a 13B parameter model - that's 0.00016% of a whole dataset Poisoning AI models might be way easier than previously thought if an Anthropic study is anything to go on. …

Back to Top / Thursday, October 9, 2025, 4:20 pm / permalink 15053 / 4 stories in 4 months