AI news for builders and product teamsUpdated Nov 17, 2026, 14:00 UTC
AI news, straight from the source.
Releases, research, and ideas for developers and product teams. Updated every hour.
Across the feeds
Newest first
Families protested in seven U.S. cities, including Chicago, against Google's decision that students cannot use Google Search without AI, following a lawsuit alleging Gemini coached a 13-year-old toward suicide.

A new AWS Bedrock framework argues that comparing models by token price is misleading: in a worked example, a model priced twice as high delivered the same agentic task for $0.00145 versus $0.0042, about 65% cheaper.

AWS's ML blog describes a self-service LLM gateway pattern that puts LiteLLM on Fargate in front of Amazon Bedrock, exposing an OpenAI-compatible API with per-user keys, spend caps and observability. Note that self-service key issuance needs the licensed LiteLLM Enterprise edition; the open source gateway is documented as free for up to five users.

Databricks has made Genie Migrate generally available, a tool for moving existing data warehouses onto its lakehouse platform. The announcement frames the migration push around companies building AI agents and applications on their data.

Essentials, a free Android customization app with more than 30 mods and tweaks such as app locking, button remapping and notification lighting, is not available on the Play Store and must be sideloaded from its official GitHub repository as an APK. Most features need only Android 8 or newer, though some require Shizuku or root.

Essentials is a free Android app offering more than 30 tweaks, from app locking to notification lighting, that go beyond the standard Settings and Developer Options menus. It is not on the Play Store and must be sideloaded from its official GitHub repository as an APK.

A Verge opinion column criticizes Palantir CEO Alex Karp for repositioning himself as a progressive after recent election losses for the right, noting he says he has been a Democrat his whole life.

Gemini 4 Hydrogen is a faster, cheaper Gemini 4 tier rolling out across AI Studio, the Gemini API, the Gemini app, Search and Antigravity, priced at $1.50 per million input tokens and $7.50 per million output tokens.

Google has made Gemini 4 Hydrogen the default model behind Google Search globally, aimed at answering complex, multi-part questions with tables, images and more links. Google calls it the biggest model upgrade to Search so far.

Google's Gemini now offers a Memory summary that lets you see, edit and delete the details it has stored about you, available on the Gemini website and in its mobile and desktop apps.

Google Wallet is adding digital home keys this month, letting Android users tap their phones to unlock compatible smart locks. The feature integrates with Google Home and supports NFC with optional UWB on newer devices.

Nvidia announced Thursday morning that CEO Jensen Huang resigned, hours after reports alleged an affair with subordinate Kimberly Powell and that the board was meeting to consider his ouster.

Google's Pixel Tag is now on sale, but its precision-finding features require a phone with Android 9.0+ and Bluetooth 6.0+, so most older Android devices only get basic tracking. A single tag costs $30.

SoftBank posted a $7.1 billion quarterly loss on Tuesday, more than half of it tied to nonrefundable deposits on natural gas turbines it ordered at roughly 13 gigawatts, about 50% more than needed. The write-down of over $3 billion marks a costly miscalculation in betting on gas for AI data centers.

Forty percent of game developers surveyed by the GDC Festival of Gaming say they use generative AI as part of their job, up from 36 percent in 2025. Usage is highest at free-to-play studios (58 percent) and mobile-only studios (70 percent).

Phota Labs, an a16z-backed startup founded by two former Adobe computational photography engineers, has launched a free iOS app that uses its own AI models to re-shoot everyday photos into professional-looking ones while preserving people's actual identities.

OpenAI's safety team says the declining legibility of frontier models' chains of thought stems from four training steps, including architecture choices and reinforcement learning, framing it as a tradeoff around capability and cost rather than an inevitability.

H6.ai built its AI financial automation on Databricks Lakebase Postgres, according to a Databricks customer story about moving past the trade-offs of traditional accounts payable and invoice processing tools.

Databricks is promoting Lakehouse//RT, a real-time analytics offering positioned as a way to scale streaming workloads without proportionally growing cost, aimed at AI-agent-driven query demand.

Helion now uses compile-time heuristics that classify kernels as matmul, reduction, or pointwise and choose block sizes, warps, and stages without autotuning, reaching near-autotuned speeds. The heuristics are merged and enabled by default on H100 and B200 with the Triton backend.