Safari 27.3 beta can darken light-only websites without changing their CSS
Safari can now adapt light-only websites for users who choose dark mode. Authors can provide their own dark scheme, but opt-out behaviour is incomplete in the first beta.
Search
Search results for “dark mode”
Find published research by company, product, platform or technology.
Showing 1–20 of 528 dossiers
DeadTag turns abandoned website scripts into an iPhone spyware delivery network
Old third-party JavaScript includes can become a live supply-chain attack when their domains expire. DeadTag shows why removing abandoned tags and monitoring what scripts load is operational security, not housekeeping.
Google expands AI Mode ads to 10 markets, including New Zealand, and launches US follow-up ads
AI Mode ads are expanding internationally, including New Zealand, and appearing in US follow-up conversations — with aggregate reporting and no AI-only campaign switch.
GitHub Copilot Fusion goes GA with multi-model routing, effort tiers and experimental local inference
GitHub's multi-model coding orchestration is now GA, with admin-controlled model pools and per-model charges. Local routing remains experimental and cloud-dependent.
Marketing & DistributionSEO & SearchGEO & AI SearchPaid AcquisitionConversion & AnalyticsActive dossier
Google starts testing AI Mode ads for users who connect Gmail and Photos context
Google reversed its earlier ad-free approach to Personal Intelligence in AI Mode. Connected-app-derived insights may influence ads, with testing scope and advertiser controls still unclear.
Apple extends App Store moderation to AI-generated content and tightens hidden-feature rules
AI image and social-app developers face clearer moderation and review obligations; hiding features can threaten an entire developer account.
Google Preferred Sources now influences publisher visibility in AI Overviews and AI Mode
Google is moving source links higher in AI answers, testing prominent official-site panels and expanding news carousels, while Preferred Sources use has grown more than sevenfold in 2026.
Meta opens Muse Spark’s 816B weights — but large model-as-a-service operators need a separate licence
Meta Muse Spark Open is genuinely downloadable, unlike a promised future weight release. But 42B active parameters do not make its 816B weights cheap to host, and the Muse Model License is not Apache or MIT: model-as-a-service operators over $20M in affiliated annual gross revenue need a separate agreement.
Together Link puts open models behind the coding-agent harnesses developers already use
Together Link connects six existing coding-agent/desktop harnesses to open models with reversible profiles, per-session routing and cost receipts. The important shift is portability at the harness boundary, not Together's unverified savings claim.
AWS Strands Decider 2B turns bounded agent decisions into an open local model
Jev made bounded decision models visible; Strands Decider makes the pattern reproducible inside an agent stack. AWS replaced Qwen3.5-2B's language-generation head with a small scoring head and released the recipe, creating a local alternative for decisions that do not need a full generative model.
GLiNER2.5-Decide puts structured AI decisions on a 340M CPU model
GLiNER2.5-Decide attacks the same bounded-decision layer as Jev and CLM from a much smaller encoder architecture. Its strongest benchmark claims are vendor-produced, but CPU deployment and constrained joint decoding make it a materially different option for software-facing AI decisions.
Jev becomes Vercel AI Gateway’s fastest-adopted model in its first 24 hours
Jev’s launch claims were interesting; Vercel’s usage data is more useful. Nearly 13% of paid AI Gateway teams tried the typed decision model in its first day, while Jev also rose to a material share of gateway requests. That does not establish retention or production success, but it is unusually fast developer uptake for a model designed to make bounded software decisions rather than generate prose.
Tencent Hy4 preview brings a 770B open-weight model to coding and long-context knowledge work
Hy4 preview is a very large sparse model with public full and FP8 weights, native speculative decoding and a 1M-token context path. Its open release makes Tencent’s claims testable, while the 1.56TB full checkpoint keeps self-hosting firmly in server-scale territory.
GLM-5.3-Flash turns the anonymous Ox Alpha trial into an open-weight multimodal coding model
GLM-5.3-Flash combines open weights, multimodal coding/agent capability and an 18B-active sparse architecture with a large anonymous pre-launch trial. Z.ai has already issued a chat-template correction for early downloads, showing that day-one self-hosted deployments need artifact-level validation as well as model benchmarking.
Anthropic’s Model Hardware Standard gives AI agents a shared interface for physical devices
MHS is an attempt to make microscopes, liquid handlers, robotic arms and other programmable hardware look like a consistent tool surface to AI agents. It is still a research preview, but the interoperability layer is already being tested with research institutions and hardware vendors.
Node.js July security release fixes high-severity HTTP/2 and Permission Model flaws
Node.js shipped v22.23.2, v24.18.1 and v26.5.1 to close a set of runtime vulnerabilities including an HTTP/2 use-after-free and a Permission Model path-matching bug that can over-grant filesystem access.
Meta’s Muse Glimmer puts a 30B multimodal agent model on 24GB-class hardware
Meta’s Muse Glimmer 30B combines tool use, coding, vision and agentic task completion with official local-runtime artifacts. A 17GB GGUF build targets 24GB-VRAM machines, but Meta also attaches a separate usage policy, so builders should distinguish weight availability from unrestricted use.
Raspberry Pi 5's 16GB model climbs to $385 as memory shortages reshape edge-compute costs
The latest DRAM-driven increase forces makers, edge deployments and small teams to re-evaluate whether high-RAM single-board computers still offer the right economics.
Google Merchant API exposes AI Mode and Gemini checkout funnels — with an alpha reporting trap
Retailers and commerce-tool developers can now query first-party UCP shopping funnels rather than infer agentic checkout performance from traffic alone, but the alpha API needs careful query design.
Mistral Large 4 weights are downloadable — but hosted model services need a separate licence
The trillion-parameter Mistral Large 4 is no longer API-only; builders can self-host, but must assess licence and hardware limits.