
SSPAI Morning Brief: Google I/O 2026: Google Launches Gemini Omni, AI Agents, and Next-Generation AI Search Features
Morning Brief
- Google announces multiple updates across its Google Gemini product lineup
- LG Electronics unveils new UltraGear gaming monitors
- Andrej Karpathy joins Anthropic
- CHERRY XTRFY launches the K5 Ultra keyboard
- Sony raises prices for select PlayStation Plus subscription plans
- Apple previews multiple accessibility feature updates powered by Apple Intelligenc
- Google adds YouTube Premium Lite benefits to its AI Pro subscription plan
- News Worth a Quick Look
Google announces multiple updates across its Google Gemini product lineup
In the early hours of May 20, Google I/O 2026 officially opened. During the keynote presentation, Sundar Pichai unveiled a wide range of product updates centered around Google Gemini.
Google announced the launch of the Gemini Omni series, a new family of models focused on video generation and editing, with the first model being Gemini Omni Flash. Omni supports using images, text, video, or audio as reference inputs and enables multi-turn video editing through natural language. Throughout the editing process, the system maintains character consistency, physical logic, and scene context. Combined with Gemini’s real-world knowledge capabilities, Omni can generate outputs that better adhere to the laws of gravity, momentum, fluid dynamics, and other physical principles. Gemini Omni Flash is rolling out immediately to global Google AI Plus, Pro, and Ultra subscribers through the Gemini app and Google Flow. Starting this week, it will also become freely available inside YouTube Shorts and YouTube Create. API access for developers and enterprise customers will launch over the coming weeks. Source
Google also officially launched Gemini 3.5 Flash, the first model in the Gemini 3.5 family, focused on agentic workflows and programming tasks. The model is now fully deployed across the Gemini app, Google Search AI Mode, Google Antigravity, the Gemini API, and related platforms and services. Gemini 3.5 Flash reportedly achieved scores of 76.2% on Terminal-Bench 2.1, 1656 Elo on GDPval-AA, 83.6% on MCP Atlas, and 84.2% on CharXiv Reasoning, while generating output tokens at four times the speed of other frontier models. The model also supports collaborative sub-agent execution through the new Antigravity harness, enabling complex multi-step workflows and programming tasks. It has already become the default model powering both the Gemini app and AI Mode inside Google Search. Google stated that the Gemini 3.5 series was developed under its Frontier Safety Framework, with enhanced cybersecurity and CBRN protections. Gemini 3.5 Pro has already entered internal testing and is scheduled for release next month. Source

The Gemini app itself is also receiving multiple updates. In addition to built-in support for Gemini 3.5 Flash and the Gemini Omni video model, the app introduces a new Neural Expressive design language, personalized morning summaries called Daily Brief, a full-time personal AI agent named Gemini Spark focused on task execution, and a new macOS client integrating Spark alongside upgraded voice features. After receiving user permission, Daily Brief can connect to Gmail and Calendar to automatically organize urgent emails, schedules, and follow-up tasks while recommending actions based on personal priorities. Gemini Spark, built on Gemini 3.5 and the Antigravity harness, can connect to Workspace tools including Gmail, Docs, and Slides. Initial MCP integrations include Canva, OpenTable, and Instacart, with support for custom sub-agents and local browser interactions planned later. Daily Brief is rolling out immediately to paid Google AI subscribers in the United States, while Gemini Spark begins limited testing this week before expanding to Google AI Ultra users in the U.S. next week. Spark functionality updates and the upgraded voice system for the Gemini macOS client are expected to launch this summer. Source

Google Search is also receiving a large wave of AI updates. A new intelligent search box can now dynamically expand based on search keywords and supports text, images, files, videos, and even Chrome tabs as input sources. The feature is now rolling out in all countries and regions where AI Mode is available. AI Overviews additionally now support follow-up questioning and extended AI conversations. Google Search is also introducing an “information agent” capable of continuously monitoring blogs, news sources, social posts, and real-time financial, shopping, and sports data in the background. This feature is expected to launch first for Google AI Pro and Ultra subscribers this summer. Additional capabilities — including local experience bookings, service reservations, and AI-powered phone calls to businesses on behalf of users — are also planned for U.S. rollout this summer. Google further previewed Antigravity-powered custom mini apps capable of generating interactive search experiences tailored to specific search queries, initially launching for U.S.-based Google AI Pro and Ultra users over the coming months. Finally, AI Mode officially announced support for personalized AI services across 98 languages in 200 countries and regions, with no subscription required. The service can connect to Gmail and Google Photos to provide personalized search results, with Google Calendar integration planned later. Source

Google also introduced Universal Cart, a cross-service and cross-merchant intelligent shopping cart system. Users will be able to add products directly through Search, Gemini, YouTube, Gmail, and other Google products, while Gemini models track discounts, price drops, pricing history, and restock alerts in the background. Built on top of Google Wallet, Universal Cart can recognize payment benefits, membership information, and merchant discounts while reasoning across multiple retailers to infer user needs. Checkout is powered by the Universal Commerce Protocol (UCP), allowing users to complete purchases using Google Pay across brands including Nike, Sephora, Target, Ulta Beauty, Walmart, Wayfair, and Shopify merchants such as Fenty and Steve Madden, or alternatively redirecting users to merchant websites for final checkout while brands remain the official sellers of record. Universal Cart will launch this summer in U.S. Search and Gemini services, with YouTube and Gmail integrations arriving later. Source

Finally, Google also shared new progress updates regarding Android XR smart glasses, Workspace AI updates, expanded support for SynthID digital watermarking technology, new Google Flow updates, and Wear OS 7.


LG Electronics unveils new UltraGear gaming monitors
On May 20, LG Electronics announced a new 24.5-inch UltraGear gaming monitor, describing it as the world’s first Full HD gaming monitor with a native 1000Hz refresh rate. The display features a 1920×1080 resolution, uses an IPS panel with a low-reflection coating, and is capable of delivering a 1000Hz refresh rate at Full HD resolution. The monitor includes a compact stand with tilt and height adjustment support, along with an integrated headphone hook. On the software side, it supports on-device AI scene optimization capable of automatically adjusting image settings based on game genres, while also offering AI Sound spatial audio features. LG has not yet announced pricing or release timing for the monitor. Source

Andrej Karpathy joins Anthropic
On May 19, renowned AI researcher Andrej Karpathy announced that he has joined Anthropic, returning once again to frontier large language model research. Karpathy was one of OpenAI’s co-founders. After leaving OpenAI in 2017, he joined Tesla to lead autonomous driving initiatives, departed in 2022, returned to OpenAI in 2023, and left again in 2024 to establish the AI education startup Eureka Labs.
Starting this week, Karpathy has joined Anthropic’s pretraining team under the leadership of Nick Joseph. The pretraining division is responsible for large-scale training of Claude’s core knowledge and capabilities and represents one of the most compute-intensive and expensive stages of frontier model development. Anthropic stated that Karpathy will build a team focused on accelerating pretraining research using Claude itself. Karpathy said he believes the coming years will be a critical phase for frontier LLM development, while also emphasizing that he remains passionate about education and plans to return to related projects in the future. Source
CHERRY XTRFY launches the K5 Ultra keyboard
CHERRY officially launched the new wired magnetic-switch keyboard CHERRY XTRFY K5 Ultra on May 19. The CHERRY XTRFY K5 Ultra features the new MK Crystal Magnetic Switch Plus and uses a compact 65% layout. Specifications include an 8kHz polling rate, a 16kHz full-key scanning rate, and adjustable Rapid Trigger distances ranging from 0.1mm to 3.3mm with 0.01mm precision. Internally, the keyboard uses an aluminum alloy positioning plate and a four-layer sound-dampening structure. The PCB is also compatible with certain third-party magnetic switches. The keyboard is priced at 499 yuan. Source

Sony raises prices for select PlayStation Plus subscription plans
On May 19, Sony announced price increases for short-term PlayStation Plus subscriptions for new users in certain regions. Monthly subscriptions have been adjusted to $10.99, while quarterly subscriptions now cost $27.99. Sony stated that the price increases were driven by “market conditions.” Existing subscribers are not immediately affected by the changes, though current users who modify or interrupt their subscriptions in the future will also be charged under the new pricing structure. Pricing for annual subscriptions and the Extra and Premium tiers remains unchanged for now. Source
Notably, some PlayStation users have already filed a class-action lawsuit against Sony Interactive Entertainment in the U.S. District Court for the Northern District of California. The lawsuit alleges that after tariffs imposed under the Trump administration’s use of the International Emergency Economic Powers Act in 2025 led Sony to raise PS5 pricing and pass costs on to consumers, the company continued retaining the benefits of those price increases even after the U.S. Supreme Court ruled 6–3 that the tariff authorization was invalid and companies were eligible for government refunds. Plaintiffs described the situation as creating a “double recovery windfall” for Sony. Source
Apple previews multiple accessibility feature updates powered by Apple Intelligenc
On May 19, Apple previewed a range of new accessibility updates powered by Apple Intelligence, covering VoiceOver, Magnifier, Voice Control, Accessibility Reader, video captions, and Apple Vision Pro. Among the updates, VoiceOver’s new Image Explorer feature can provide far more detailed system-wide descriptions of photos, bills, personal records, and other image-based content. Magnifier now supports invoking Live Recognition through the iPhone Action Button to identify objects within the camera view and answer follow-up questions, while also supporting voice commands such as “zoom in” or “turn on flashlight.” Voice Control gains support for natural-language screen interactions, eliminating the need to memorize button labels or numbered overlays. Accessibility Reader can process complex documents containing multiple columns, images, and tables while preserving customized fonts and color settings. Device-side generated captions will automatically create subtitles for videos that do not already include caption tracks, supporting iPhone, iPad, Mac, Apple TV, and Apple Vision Pro. Vision Pro will additionally support eye-tracking control for compatible powered wheelchairs, initially supporting Tolt and LUCI alternative driving systems in the United States through both Bluetooth and wired connections. Name Recognition functionality is expanding to support 50 languages, tvOS will introduce larger text support, and Made for iPhone hearing aids are receiving improved multi-device handoff functionality. Source

Google adds YouTube Premium Lite benefits to its AI Pro subscription plan
On May 19, Google announced that YouTube Premium Lite will now be included as part of the Google AI Pro subscription package, while the full YouTube Premium plan will become part of the Google AI Ultra subscription tier. Google AI Pro individual subscribers will receive a personal YouTube Premium Lite subscription, including ad-free playback for select videos, offline downloads, and background playback functionality.
Google AI Pro currently costs $20 per month and already includes a Gemini subscription, 5TB of shared storage across Gmail, Google Drive, and Google Photos, alongside YouTube Premium Lite access. Google AI Ultra is currently priced at $100 per month and includes the full individual YouTube Premium plan. Existing YouTube Premium and Premium Lite subscribers will not have their subscriptions automatically canceled and must manually cancel their current subscriptions before activating YouTube benefits through the AI subscription plans. Users currently on AI subscription trial periods must also convert to full paid subscriptions before receiving the included YouTube benefits. For family subscription plans, only the family administrator is eligible for the bundled benefits, which cannot be shared with other family members. Source
News Worth a Quick Look
- On May 19, Google rolled out updated app icons across its iOS applications and web services, including changes to the web-based Google app switcher and several Google apps on iOS. Source
- X has further restricted the daily posting and reply limits for non-paying users. Free unverified accounts are now limited to 50 original posts and 200 replies per day, down from the previous 2,400 daily posts. X stated that the changes are intended to combat spam and bot activity on the platform. Source
- Mozilla released Firefox 151, introducing generative AI control toggles on both iOS and Android versions of the browser. Users can now disable all generative AI features with a single switch or individually enable and disable specific AI capabilities. Source
- Plex announced that beginning in late June 2026, the one-time lifetime purchase price for Plex Pass will increase from $250 to $750 — a 200% increase. Existing lifetime subscribers will not be affected. Plex Pass originally launched in 2012 at $75 and has undergone multiple price increases over the years, including a jump to $250 in March 2025. Plex stated that monthly and annual subscription models are more sustainable for long-term development, and while the company previously considered eliminating the lifetime option entirely, it ultimately decided to keep it available at a significantly higher price. Source
- On May 19, Microsoft confirmed that, following enterprise users, SMS verification codes for personal accounts will also gradually be phased out. Microsoft stated that SMS verification transmits data in plaintext and carries risks of interception. The company is now aggressively promoting passkeys as the default login method for personal accounts. Passkeys are based on the FIDO2 standard and use public-key cryptography, with private keys stored securely inside the TPM security chip on user devices and never transmitted across networks. Source
Leave a Reply