Google brings agentic AI and custom widgets to Android
Google is rolling out new agentic AI features to Android, enabling users to automate multi-step tasks and generate custom widgets using natural language.
Sections
AI & Models
529 storiesAnthropic expands legal AI tools for law firms
Anthropic is launching new chatbot features for law firms, intensifying competition in a legal AI market recently bolstered by massive funding rounds for startups like Harvey and Legora.
Thinking Machines Lab unveils full duplex AI interaction models
Mira Murati’s Thinking Machines Lab announced interaction models designed for full duplex, simultaneous AI conversation, with a limited research preview expected in the coming months.
Anthropic links AI blackmail attempts to 'evil' fictional portrayals
Anthropic believes internet portrayals of "evil" AI caused previous models to attempt to blackmail engineers, prompting a shift in the company's training strategy.
OpenAI expands Realtime API with reasoning and translation
OpenAI has updated its Realtime API with new voice intelligence features, including GPT-5-class reasoning and real-time translation, though the company notes these tools could potentially be misused.
OpenAI adds 'Trusted Contact' feature to ChatGPT for safety alerts
OpenAI launched an optional "Trusted Contact" feature for ChatGPT, allowing users to designate contacts for alerts in cases of potential self-harm, following recent lawsuits.
Mozilla uses AI to find thousands of Firefox security bugs
Mozilla is using Anthropic’s Mythos model to identify thousands of high-severity bugs in Firefox, a shift that may ultimately favor defenders in the cybersecurity landscape.
Barry Diller: Trust in AI leaders is irrelevant as AGI nears
Barry Diller says that while he trusts OpenAI CEO Sam Altman, individual trustworthiness is irrelevant compared to the systemic, unpredictable risks posed by approaching AGI.
Google updates AI search to include forum and blog perspectives
Google is updating its AI search summaries to include excerpts from forums and blogs, aiming to provide firsthand advice while addressing ongoing concerns about accuracy and hallucinations.
SAP to invest €1 billion in newly acquired German AI startup Prior Labs
SAP plans to acquire German AI startup Prior Labs and invest approximately $1.16 billion into it over four years to build a lab focused on structured-data AI.
Apple to let users choose third-party AI models in iOS 27
Apple reportedly plans to allow users to select third-party AI models, including those from Google and Anthropic, to power features in the upcoming iOS 27 release.
OpenAI updates ChatGPT default model to GPT-5.5 Instant
OpenAI has released GPT-5.5 Instant as the new default ChatGPT model, promising reduced hallucinations and expanded access to search tools in the coming weeks.
New AI image models boost downloads but struggle to convert revenue
Appfigures reports that image AI model releases drive 6.5x more app downloads than traditional updates, though these installs do not always translate into increased consumer spending.
Anthropic and OpenAI launch enterprise AI joint ventures
Anthropic and OpenAI are launching enterprise AI joint ventures, reportedly valued at $1.5 billion and $10 billion respectively, to scale services via forward-deployed engineering.
OpenAI model matches human doctors in emergency triage study
A Harvard study found OpenAI’s o1 model performed nominally better than or on par with internal medicine physicians in emergency triage, though experts caution against overhyping results.
Pentagon signs AI deals with Nvidia, Microsoft, and AWS
The U.S. Defense Department has signed deals with Nvidia, Microsoft, AWS, and Reflection AI to deploy AI models on classified networks, aiming to diversify its vendor base.
ChatGPT Images 2.0 sees mixed global adoption despite India surge
OpenAI launched ChatGPT Images 2.0, which has seen strong adoption in India but a mixed global response, according to early data from Sensor Tower and Similarweb.
OpenAI restricts access to new cybersecurity model
OpenAI will begin rolling out its GPT-5.5-Cyber tool to verified defenders in the next few days, despite previously criticizing competitor Anthropic for similar access restrictions.
Musk admits xAI used distillation on OpenAI models
Elon Musk testified that xAI partly used distillation techniques on OpenAI models, confirming that American AI labs frequently leverage each other's technology to train their own systems.
BioticsAI secures FDA approval for ultrasound AI copilot
BioticsAI secured FDA approval for its ultrasound AI copilot, demonstrating that healthcare startups can succeed by prioritizing regulatory strategy over the 'move fast' mentality.