Last Updated on May 14, 2025 by Editor
AI Week # 19. Here’s What the Future Just Unfolded—And You Can’t Afford to Stay Behind.
From multi-million-dollar AI funding rounds to regulatory U-turns, and from Google’s AI landing on your wrist to Spotify letting you talk to your DJ, the week of May 7 to May 13, 2025, was nothing short of groundbreaking. Major players like Microsoft, Netflix, Meta, TikTok, Cisco, and Google rolled out updates, tools, and deals that are reshaping AI’s global impact across entertainment, security, development, and policy. While some startups emerged from stealth with tens of millions in backing, others launched tools to fight AI hallucinations or boost cybersecurity at scale. Meanwhile, the U.S. government quietly shifted gears on export restrictions, signaling deep geopolitical implications. Whether you’re an AI enthusiast, tech investor, or digital trend-watcher, this recap is your essential pulse check on where AI is heading next—and why you can’t afford to tune out.
Saudi Arabia Unveils $600B AI and Defense Investment Package
On May 13, 2025, Saudi Crown Prince Mohammed bin Salman announced a monumental $600 billion investment package focusing on artificial intelligence (AI) and defense sectors. This initiative includes the establishment of a new AI company, Humain, funded by the Public Investment Fund (PIF), which manages assets worth approximately $940 billion. The announcement coincided with the U.S.-Saudi Investment Forum in Riyadh, attended by prominent figures such as Donald Trump, Elon Musk, Sam Altman, and Mark Zuckerberg.
Key components of the package involve a $142 billion arms deal to enhance Saudi Arabia’s military capabilities, and significant investments in AI infrastructure, including data centers. Notably, DataVolt, a Saudi firm, plans to invest $20 billion in U.S. data centers and energy infrastructure. Additionally, U.S. tech giants like Google and Oracle have pledged $80 billion in joint ventures. These strategic moves underscore Saudi Arabia’s ambition to become a global AI hub and strengthen U.S.-Gulf economic ties.
US Revamps AI Chip Export Controls, Scraps “AI Diffusion Rule”
Washington D.C. – The United States government has overhauled its approach to regulating the export of advanced Artificial Intelligence (AI) chips, announcing the rescission of the controversial “AI Diffusion Rule” just days before it was set to take effect on May 15, 2025. This move by the Trump administration aims to foster American innovation and strengthen diplomatic ties with allied nations, while still focusing on preventing advanced technology from reaching adversaries.
The Biden-era “AI Diffusion Rule” had proposed a three-tiered system for AI chip exports, which was met with criticism from U.S. tech companies like Nvidia and AMD, as well as international partners. These entities argued that the rule was overly complex, would stifle innovation, and could potentially harm U.S. competitiveness. The Commerce Department stated the rule “would have undermined U.S. diplomatic relations with dozens of countries by downgrading them to second-tier status.” The reversal is seen as a positive development for chip manufacturers, with Nvidia’s stock reportedly surging following the announcement. While the “AI Diffusion Rule” is scrapped, the Commerce Department emphasized that strict export controls targeting China and other adversarial nations remain in place and will be robustly enforced. New guidance includes clarifying that the use of Huawei’s Ascend AI chips globally violates U.S. export controls and warning U.S. companies about protecting supply chains from diversion tactics. The administration plans to develop a replacement rule in the future, focusing on a “bold, inclusive strategy to American AI technology with trusted foreign countries.”
Cisco Deepens Saudi Arabia AI Partnership to Fuel Vision 2030
Riyadh, Saudi Arabia – Cisco has announced a significant expansion of its partnership with Saudi Arabia, joining forces with the Kingdom’s new AI enterprise, HUMAIN, to build out its AI infrastructure. This collaboration is a cornerstone of Saudi Arabia’s Vision 2030, which aims to diversify its economy and establish the nation as a global leader in digital innovation.
The multi-year initiative will leverage Cisco’s cutting-edge technologies, including its Nexus portfolio, HyperFabric, UCS systems, and security solutions like Hypershield and Splunk, to create a scalable, secure, and resilient AI infrastructure. Beyond technology deployment, the partnership focuses heavily on talent development and research. Cisco plans to establish a Cisco AI Institute at King Abdullah University of Science and Technology (KAUST) and has committed to providing free training in AI, cybersecurity, data science, and programming to 500,000 learners over the next five years through its Networking Academy. This builds on Cisco’s 25-year presence in the Kingdom and its existing initiatives, such as operationalizing data centers for cloud security and the Webex Collaboration Platform, and plans for a Meraki cloud region. The collaboration also extends to supporting digital infrastructure for major upcoming events like Riyadh Expo 2030 and the 2034 FIFA World Cup. This move is part of Cisco’s broader strategy in the region and its global AI focus, which includes a $1 billion Global AI Investment Fund.
Google Fortifies Android 16 with “Advanced Protection” and “Lockdown” Features for At-Risk Users
Google is significantly enhancing mobile security with the introduction of an expanded Advanced Protection Program integrated at the device level in Android 16. This initiative, which includes features comparable to a “Lockdown” mode, is designed to provide Google’s strongest security measures for individuals at high risk of targeted online attacks, such as journalists, activists, and political figures.
The revamped Advanced Protection acts as a unified control center, activating and locking Android’s most robust security settings across the device and integrated Google apps like Chrome, Messages, and Phone. A key innovation is “Intrusion Logging,” an industry-first feature that securely backs up device logs in a privacy-preserving, tamper-resistant manner, accessible only to the user for forensic analysis if a compromise is suspected. Other critical enhancements include USB port lockdown (turning connections into charging-only by default when the device is locked), options to disable auto-reconnection to insecure Wi-Fi networks, and further integration with Scam Detection for Phone by Google. While core enhanced security features will be available with the initial release of Android 16, advanced capabilities like Intrusion Logging and enhanced USB protection are slated to roll out later in the year. This move underscores Google’s commitment to a “defense-in-depth” strategy, aiming to protect users from sophisticated cyber threats by making top-tier security more easily accessible.
Spotify’s AI DJ Spins a New Tune with Interactive Voice Commands
Spotify’s innovative AI DJ feature is receiving a significant upgrade, now allowing Premium subscribers to personalize their listening experience using voice commands. This development, rolling out in over 60 markets, empowers users to verbally request specific artists, genres, moods, or even ask for music suitable for particular activities. The enhancement aims to make music discovery more interactive and tailored, moving beyond the AI DJ’s initial offering of curated playlists with AI-driven commentary. User engagement with the AI DJ has reportedly almost doubled in the past year, indicating a strong appetite for AI-enhanced music experiences.
The AI DJ, voiced by Xavier “X” Jernigan, Spotify’s Head of Cultural Partnerships, has been a unique offering since its launch, blending personalization technology with generative AI to create a radio-like feel. While this new voice interaction adds a layer of convenience and deeper personalization, some users note potential awkwardness in using voice commands in public or quiet environments. Currently, voice input is the primary method for these new interactions. This update signifies Spotify’s continued investment in AI to create a more dynamic and responsive music platform, further differentiating its service in the competitive streaming landscape.
Google’s Gemini AI Expanding its Reach to Wear OS Smartwatches and Google TV
Google is significantly broadening the availability of its advanced AI model, Gemini, bringing its capabilities to Wear OS smartwatches in the coming months and to Google TV later this year. This strategic expansion aims to embed Gemini more deeply into the everyday lives of users, offering a more intuitive and conversational AI assistant across a wider range of devices. For Wear OS users, Gemini will enable natural language interactions for tasks like setting reminders, sending messages, or getting information directly from their wrist, even when their phone isn’t readily accessible. This move is poised to enhance the utility of smartwatches beyond basic notifications and fitness tracking.
On Google TV, Gemini will transform the content discovery experience by providing personalized recommendations based on natural language queries. Users will be able to ask for specific types of movies suitable for children, explore topics with linked YouTube educational content, or get more information about what they’re watching. The rollout for Google TV will reportedly begin with upcoming TCL television sets before becoming available on Google’s own streaming devices. This integration underscores Google’s ambition to create a unified AI experience across its ecosystem, including Android Auto and future XR platforms, making Gemini a central component of user interaction with Google’s services and hardware.
TikTok Unveils ‘AI Alive’: A New Tool to Animate Static Images into Videos
TikTok has launched a new generative AI feature named “AI Alive,” designed to transform static photos into dynamic short videos within TikTok Stories. This innovative tool allows users to upload an image via the Story Camera and then use text prompts to describe how they want the AI to animate it, adding movement and various creative effects. The introduction of AI Alive marks TikTok’s latest step into the burgeoning field of AI-powered content creation, providing users with new ways to express themselves and enhance their storytelling on the platform.
Recognizing the importance of responsible AI deployment, TikTok is implementing several safety measures for “AI Alive.” Content generated by the tool will be watermarked with an “AI-generated” label and include C2PA (Coalition for Content Provenance and Authenticity) metadata to ensure transparency about its origin. Furthermore, both the uploaded image, the user’s text prompt, and the final AI-generated video will undergo moderation checks before and after publishing. While the tool offers exciting creative possibilities, some early user reports suggest the AI-generated animations can sometimes produce quirky or unexpected results. This launch positions TikTok competitively alongside other social media giants who are also exploring and integrating AI content generation tools.
AllTrails Unveils “Peak” Membership with Trailblazing AI Route Planning
AllTrails is revolutionizing outdoor adventure planning with the launch of its new premium membership tier, “Peak,” priced at approximately $80 per year (or $79.99 / £79.99). This offering, more than double the cost of its popular “Plus” tier, introduces cutting-edge AI-powered features designed to enhance trail discovery and safety. The “Peak” tier, which began its global rollout with early access in May 2025 and aims for full availability by June 2025, signifies a major upgrade to the platform that was named Apple’s iPhone App of the Year in 2023. AllTrails, which secured $150 million in funding in 2021, is leveraging AI to stay competitive and cater to its loyal user base.
The cornerstone of the “Peak” membership includes AI-driven custom route creation, allowing users to generate new trails or modify existing ones based on preferences like desired length, reduced steepness, or enhanced scenic beauty. Furthermore, subscribers will benefit from real-time and forecasted trail conditions, aggregating data from 15 different weather factors including temperature, precipitation, snow depth, air quality, and even mosquito activity forecasts. Other premium features include community heatmaps to gauge trail traffic and an upcoming “Outdoor Lens” feature (expected later in the summer) that will use a phone’s camera and AI to identify flora, fauna, and fungi. All AllTrails tiers will also now feature over 500,000 Points of Interest.
AI Search Unicorn Perplexity AI Eyes Massive $500M Funding, Valuation to Hit $14 Billion
AI search startup Perplexity is reportedly in advanced discussions to secure a new funding round of approximately $500 million, which would catapult its valuation to an impressive $14 billion. This potential investment, with Accel rumored to be a leading investor, underscores the explosive growth and investor confidence in the generative AI search sector. Perplexity, known for its AI-powered search tools that provide summarized answers with cited sources, has seen a meteoric rise in valuation, from $520 million in January 2024 to a reported $3 billion in a previous round and $9 billion in December 2024. This new funding would significantly bolster its position against competitors.
The company, which has attracted investment from notable backers including Nvidia, Jeff Bezos (through Bezos Expeditions), IVP, NEA, SoftBank Vision Fund, and Databricks Ventures, is experiencing rapid user adoption and has reportedly reached $100 million in annual recurring revenue (ARR). Amidst this growth, Perplexity is also exploring strategic expansions, including the potential launch of its own web browser, codenamed “Comet.” There is also industry speculation about a possible integration of Perplexity’s AI search capabilities into Apple’s Safari browser, highlighting the significant impact the startup is having on the future of information discovery.
Microsoft Build 2025: A Deep Dive into AI with Major Copilot and Azure Upgrades Expected
Microsoft’s annual developer conference, Build 2025 (May 19-22), is poised to be heavily dominated by Artificial Intelligence, with significant upgrades anticipated for its AI assistant, Copilot, and its cloud platform, Azure. CEO Satya Nadella and CTO Kevin Scott are expected to lead keynotes, setting the stage for how AI will continue to reshape every layer of Microsoft’s tech stack, from developer tools to end-user experiences in Windows and Microsoft 365. The event will emphasize practical AI implementation, moving beyond theory to showcase real-world applications and tools for developers.
Key expectations for Copilot include the expansion of “Copilot Vision” to desktop environments, enabling the AI to understand and interact with on-screen content. Attendees also anticipate further details on AI Agents within Windows 11, designed to automate tasks, and enhancements to the “Researcher” tool in Copilot for advanced data gathering. On the Azure front, major updates are expected for Azure AI Foundry, focusing on agent development, model fine-tuning (including for new GPT models and third-party offerings), and multimodal capabilities. Azure AI Search is also likely to see advancements for creating sophisticated RAG (Retrieval Augmented Generation) systems. While software will be the main focus, with updates to .NET, GitHub, and Visual Studio, there have been some industry whispers about potential new, more affordable Surface hardware powered by Snapdragon X chips.
Theom Raises $20M Series A to Revolutionize Data Security with AI-Native Operations Center
Data security innovator Theom has successfully closed a $20 million Series A funding round led by Wing VC. The round also saw participation from strategic investors including SentinelOne (S Ventures), Snowflake (Snowflake Ventures), and Databricks (Databricks Ventures). This latest injection of capital brings Theom’s total funding to over $36 million, following a reported $16.4 million seed round in late 2022, which was led by Ridge Ventures and M12 (Microsoft’s Venture Fund). The fresh funds are earmarked to accelerate product development, scale go-to-market operations, and expand Theom’s engineering and sales teams, further solidifying its position in the rapidly evolving data security landscape.
Theom is making waves with its AI-native Data Operations Center (DOC) platform, designed to automate AI and data governance and security in today’s dynamic, multi-cloud environments. As businesses increasingly leverage AI, particularly Generative AI (Gen AI), the attack surface for sensitive data expands, making traditional security approaches inadequate. Theom addresses this by treating data as the new perimeter, providing real-time visibility into data identity, usage patterns, and potential risks. Its platform offers AI-powered prioritization of security threats and integrates seamlessly with major data platforms like Snowflake and Databricks, cloud providers such as AWS and Google Cloud, and various SaaS applications, including Microsoft 365 and Okta. This comprehensive approach enables organizations like FiServ, Grammarly, and JetBlue to protect their critical data assets effectively.
Google Launches $100M Fund to Back AI Startups Across the Globe
Google has announced a new initiative called “Google AI Startup Support Program”, part of a broader effort to fund and mentor early-stage companies building on Google Cloud and Vertex AI. Through this program, Google will allocate $100 million to AI startups across verticals, including healthcare, finance, sustainability, and productivity.
Launched under Google for Startups, this initiative will provide cloud credits (up to $350,000 per startup), technical mentorship, access to Tensor Processing Units (TPUs), and early access to Google’s generative AI models, including Gemini. Google will also help select startups with go-to-market support, investor matchmaking, and potential acquisition paths through its global VC ecosystem.
This comes as part of Google’s broader push to compete with Microsoft and Amazon in the enterprise AI infrastructure market, while simultaneously fostering innovation from the ground up. Startups selected for the initiative will be evaluated based on AI readiness, societal impact, and scalability potential. The program reflects Google’s long-term bet on democratizing AI development, and offers a non-equity approach for early-stage collaboration.
Aarki Launches “Encore”: AI-Powered, Privacy-First Platform to Transform Mobile App Growth
Mobile advertising technology company Aarki has announced the launch of “Encore,” its next-generation machine learning platform designed to deliver full-funnel growth for mobile marketers within the increasingly complex privacy-first landscape. Unveiled around May 12, 2025, Encore leverages advanced AI to provide smarter user acquisition and stronger user retention for mobile applications. The platform is built to address the challenges posed by evolving privacy regulations and changes in mobile operating systems, offering SKAN 4.0-ready optimization for iOS campaigns.
Key features of Aarki’s Encore platform include dynamic targeting, which utilizes a combination of attributed, non-attributed, and real-time ad call data to create precise audience segments without relying on device IDs. Its deep neural networks enable continuous model evolution, recalibrating in real-time based on live performance signals to ensure sustained campaign growth. Encore also offers unified intelligence across both user acquisition (UA) and retargeting (RT) efforts, fostering a holistic approach to mobile growth. By integrating AI with a privacy-centric design and a unified creative framework, Aarki aims to help mobile publishers and advertisers, including prominent names like Playtika and Bagelcode, maximize their return on ad spend (ROAS) and efficiently scale their user base in a compliant manner, reportedly achieving results like a +25% uplift in D7 ROAS and an 18% decrease in Cost Per Install (CPI) for its clients.
Saudi Arabia Ignites National AI Ambitions with Multi-Billion Dollar “Humain” Venture Amidst High-Profile Tech Summit
Saudi Arabia has officially launched Humain, a landmark multi-billion dollar Artificial Intelligence (AI) company, signaling a determined push to become a global powerhouse in the rapidly evolving AI sector. The announcement, made on May 12, 2025, coincided with a high-profile U.S.-Saudi investment forum in Riyadh, attended by a constellation of tech luminaries including former U.S. President Donald Trump, Tesla and SpaceX CEO Elon Musk, OpenAI’s Sam Altman, and Meta’s Mark Zuckerberg. Humain will be chaired by Crown Prince Mohammed bin Salman and is backed by the Kingdom’s formidable $940 billion Public Investment Fund (PIF), underscoring the strategic national importance of this initiative.
Humain is envisioned as the “prime vehicle to drive the kingdom’s AI strategy.” The venture will focus on investing in and operating a wide array of AI assets, including the development of Arabic large language models (LLMs), advanced data centers, and cutting-edge cloud computing infrastructure. This initiative is a core component of Saudi Vision 2030, the Crown Prince’s ambitious blueprint to diversify the nation’s economy beyond oil, particularly as global oil prices have seen recent lows. The PIF has already demonstrated its appetite for tech investments with previous backing for giants like Uber and Lucid, as well as AI-related projects with Google Cloud and Salesforce. The launch of Humain signifies a more direct approach, aiming to build and control a domestic AI ecosystem. This move follows earlier Saudi AI commitments, such as a $40 billion AI fund announced in 2024 and a data center deal with AI startup Groq. Concurrently, Wyld VC, a new Saudi-based early-stage venture capital firm, also announced its inaugural $50 million fund focused on AI, highlighting the burgeoning private investment landscape in the region’s AI sector.
Sakana AI Unveils “Continuous Thought Machines,” A Revolutionary AI Architecture for Human-Like Reasoning
AI research startup Sakana AI has introduced a groundbreaking new AI model architecture named “Continuous Thought Machines” (CTMs), designed to enable AI models to reason and process information more like the human brain, requiring less explicit guidance. The announcement on May 12, 2025, detailed how CTMs depart from traditional AI models, particularly Transformer-based Large Language Models (LLMs), by incorporating the dimension of time and neural synchronization as core operational principles. This allows CTMs to “think” through problems step-by-step, offering a more interpretable and potentially more powerful form of reasoning by allowing neurons to make activation decisions based on short-term memory.
Sakana AI, co-founded by former Google researchers David Ha and Llion Jones (who was also a co-author of the seminal “Attention Is All You Need” paper that introduced Transformers), aims to bridge the gap between the computational efficiency of current AI and the nuanced, flexible cognition of biological brains. Early results indicate CTMs show promise in tasks requiring complex sequential reasoning, such as 2D maze navigation, where they significantly outperformed existing baselines and demonstrated generalization to larger, unseen problems. On image classification benchmarks like ImageNet-1K, CTMs exhibited rich internal dynamics and good calibration. The architecture allows for adaptive computation, meaning the model can adjust its “thinking time” based on task difficulty. While still in the research phase, with challenges like longer training times compared to feedforward models, CTMs represent a significant step towards more interpretable and biologically plausible AI. Sakana AI has raised approximately $244 million to $344 million in funding from prominent investors including Khosla Ventures, Lux Capital, New Enterprise Associates (NEA), and Nvidia, with a reported valuation of around $1.5 billion as of late 2024.
Google’s Open-Source Gemma AI Models Skyrocket Past 150 Million Downloads, Fueling Developer Innovation
Google’s family of lightweight, open-source AI models, Gemma, has achieved a remarkable milestone, surpassing 150 million downloads since its debut in February 2024. Omar Sanseviero, a developer relations engineer at Google DeepMind, also revealed that developers have created over 70,000 custom variants of Gemma on the popular AI developer platform Hugging Face. This rapid adoption underscores the growing appetite within the developer community for accessible and adaptable AI tools. The Gemma models, based on the same research and technology as the larger Gemini models, are designed for efficiency and can run on devices ranging from laptops and workstations to mobile phones.
Google has positioned Gemma as a strong competitor to other open models like Meta’s Llama, emphasizing Gemma’s superior performance on certain benchmarks related to reasoning, math, and programming skills relative to its size. The Gemma family includes models of various sizes, such as 2B, 7B, 9B, and 27B parameters, with the latest iterations, like Gemma 3, boasting multimodal capabilities (understanding text and images) and support for over 100 languages. Google has also released fine-tuned versions for specific applications, including CodeGemma for code generation and versions tailored for scientific research like drug discovery. While Gemma’s download numbers are impressive, Meta’s Llama models had reportedly reached 1.2 billion downloads by late April 2025. Both model families have faced some scrutiny over their custom licensing terms, which some developers find create uncertainty for commercial use. Nevertheless, the widespread adoption of Gemma highlights Google’s significant contribution to the open-source AI ecosystem and its commitment to empowering developers worldwide.
Ex-Synapse CEO Seeks $100M for Groundbreaking Humanoid Robotics Startup
Gagan Biyani, the former CEO of Synapse, is reportedly working on a new venture aimed at building next-generation humanoid robots and is actively seeking to raise $100 million in funding, according to insiders familiar with the matter. Though still in stealth mode, the startup is focused on leveraging advanced artificial intelligence and robotics to build highly adaptable machines capable of human-like movements and real-world interactions.
Biyani, best known for his role in edtech and finance, is now venturing into robotics, signaling a growing trend where tech entrepreneurs are entering the humanoid race, competing against the likes of Tesla’s Optimus, Figure AI, and Sanctuary AI. Sources indicate that he has already held preliminary discussions with top-tier Silicon Valley VCs, with a strong emphasis on AI-powered autonomy, edge computing, and affordable scalability. The company’s goal is to provide enterprise-grade humanoids that could be deployed in warehouses, hospitals, and eldercare settings within the next few years.
This comes amid a broader surge of investment in AI robotics, with Figure recently raising $675 million at a $2.6 billion valuation. Biyani’s move may signal the next big wave in AI-driven labor automation.
Glass Imaging Secures $20M to Enhance Digital Photography With AI
Glass Imaging, a computational photography startup, has raised $20 million in Series A funding to revolutionize digital image quality using AI-powered processing pipelines. The round was led by GV (formerly Google Ventures), with participation from Future Ventures and LDV Capital. The company aims to embed its proprietary AI algorithms into smartphone and camera sensors, enhancing clarity, dynamic range, and low-light performance far beyond traditional imaging systems.
Founded by alumni from Apple and MIT, Glass Imaging’s breakthrough technology relies on deep learning-based image reconstruction from minimal sensor data. Their software enables higher-quality photos from smaller, cheaper sensors, reducing the need for bulky, expensive camera hardware. According to the team, their AI models outperform standard imaging stacks by up to 40% in sharpness and 30% in noise reduction in low-light conditions.
Glass plans to use the funding to scale R&D and collaborate with OEM partners to integrate its technology into commercial devices by 2026. With the smartphone camera market projected to hit $58 billion by 2028, Glass Imaging is positioning itself as a key player in the AI-enhanced mobile photography revolution.
Meta Recruits Former DeepMind AI Director to Lead FAIR Research Division
Meta has appointed Dr. Koray Kavukcuoglu, former Director of Research at Google DeepMind, as the new head of its FAIR (Fundamental AI Research) division. This move marks Meta’s continued effort to advance frontier AI research and compete with global leaders like OpenAI, Anthropic, and Google DeepMind.
Dr. Kavukcuoglu, renowned for his work on AlphaGo and deep reinforcement learning, brings over a decade of experience in scaling general-purpose AI architectures. He joins Meta at a time when the company is intensifying its focus on open-weight models, such as LLaMA 3, and doubling down on efforts to build Artificial General Intelligence (AGI) in a responsible and scalable manner.
His appointment comes after Meta split its AI teams in 2023 into FAIR (long-term research) and GenAI (product-focused development). FAIR has been responsible for landmark projects in self-supervised learning, multimodal AI, and AI safety. With Kavukcuoglu at the helm, industry experts anticipate an acceleration in Meta’s open science collaborations and innovation in deep learning efficiency.
The hire underlines Meta’s strategic ambition to not only lead AI innovation but also shape the ethics and accessibility of next-gen models.
Google Launches AI Tools to Shield Chrome Users From Online Scams
Google has rolled out new AI-powered tools in Chrome to combat online scams, marking a significant enhancement in browser-level user protection. This update uses real-time AI threat detection to warn users against malicious sites and suspicious pop-ups, particularly targeting phishing, fake download buttons, and deceptive login pages.
The AI tools leverage Google Safe Browsing’s upgraded infrastructure, now capable of scanning billions of URLs per day, with a 40% faster threat detection rate compared to its earlier version. One notable feature is a machine learning model integrated into Chrome’s Side Panel, offering users safety information on visited sites in real time.
This move comes amid rising digital threats — in 2024 alone, phishing attempts rose by over 38% year-over-year, according to Google’s internal security report. The AI-backed features will be available first on Chrome desktop, with a gradual rollout to mobile platforms in the coming months. Google emphasized that user privacy remains central, with no personal browsing data being stored.
By embedding AI cybersecurity tools directly into Chrome, Google aims to set a new industry standard for safe browsing in the AI era.
Clay Hits $1.5B Valuation After Employee Tender Offer Led by Sequoia
Clay, a relationship intelligence platform, has completed an employee tender offer that values the company at $1.5 billion, led by Sequoia Capital. While not a new funding round, the deal enabled early employees and stakeholders to cash out a portion of their equity, signaling growing confidence in the company’s trajectory.
Clay uses AI and machine learning to help users manage professional relationships by consolidating data from emails, calendars, and social networks. With features such as automatic contact enrichment and relationship insights, Clay is widely adopted by VC firms, sales teams, and executives managing high-value networks.
The latest valuation reflects a steep rise from its last known value in 2022, when it reportedly raised $8 million in a Seed round led by Forerunner Ventures. Sequoia’s involvement in this secondary transaction underscores a strong belief in Clay’s potential to become a category-defining AI-powered CRM alternative.
The company is also rumored to be working on integrations with major productivity tools like Slack and Salesforce. As AI becomes essential in network management and personal CRM, Clay’s valuation jump shows that investors are betting big on AI-enhanced professional intelligence tools.
Amazon Deploys New AI Tool to Automatically Enhance Product Listings
Amazon has introduced an AI-powered tool designed to help sellers generate high-quality product listings more efficiently and accurately. The tool, embedded in Seller Central, uses generative AI to automatically suggest product titles, descriptions, bullet points, and even backend search terms based on minimal seller input.
According to Amazon, early testing showed that listings created or edited with this tool had more than 20% higher conversion rates than manually created ones. Sellers simply provide a few keywords or product features, and the AI generates optimized content aligned with Amazon’s marketplace standards and SEO best practices.
The tool is built on a fine-tuned large language model, trained specifically on Amazon’s catalog and customer behavior data. It also flags non-compliant content proactively, reducing the chances of listing errors or policy violations.
This innovation is part of Amazon’s broader effort to democratize AI for small businesses, allowing even first-time sellers to compete on content quality with established brands. With millions of active sellers and billions of product listings, Amazon’s generative AI tool could redefine how e-commerce SEO and product visibility work going forward.
Samsung’s AI Strategy Drives Record Revenue Despite Semiconductor Challenges
Samsung Electronics reported a record quarterly revenue of KRW 79.14 trillion ($55.4 billion), a 10% increase year-over-year, driven by its strategic focus on AI, particularly in its Mobile Experience (MX) Business. The Galaxy S25 series, featuring Galaxy AI, contributed significantly to this success, with the MX division generating KRW 37 trillion in revenue and KRW 4.3 trillion in operating profit, the highest in four years. Despite this success, the Device Solutions (DS) Division, which includes its semiconductor operations, faced challenges, with a 42% decline in operating profit due to lower High-Bandwidth Memory (HBM) sales and export controls. Samsung is navigating global trade uncertainties and remains committed to its AI strategy, planning to expand its AI smartphone lineup and enhance its foldable lineup with improved AI user experiences.
Instacart CEO Fidji Simo Joins OpenAI to Lead Applications
Fidji Simo, the CEO of Instacart, is set to join OpenAI as its CEO of Applications. In this new role, Simo will oversee the teams responsible for turning OpenAI’s AI research into products like ChatGPT and Sora. Simo’s move comes as OpenAI focuses on scaling its core business operations. Sam Altman, while remaining the overall CEO, will concentrate more on research, computing infrastructure, and safety. Simo, who has a strong background in leading consumer-facing products, including her previous role at Facebook, is expected to help OpenAI in its next phase of growth. Instacart will soon announce a new CEO, with Simo remaining as Chair to support the transition.
Fastino Raises $17.5M to Train AI Models on Cheap Gaming GPUs
Fastino, a Palo Alto-based AI startup, has raised $17.5 million in seed funding led by Khosla Ventures, bringing its total funding to nearly $25 million. The company specializes in creating Task-Specific Language Models (TLMs) that are intentionally small and efficient, allowing them to be trained on low-end gaming GPUs costing less than $100,000 in total.service
Fastino’s innovative approach enables the development of AI models that are both cost-effective and high-performing. These models can deliver responses significantly faster than traditional large language models, making them suitable for enterprise applications such as data redaction and document summarization. The company’s team comprises AI researchers from leading institutions, including Google DeepMind, Stanford, Carnegie Mellon, and Apple Intelligence, reflecting its strong technical foundation.
ServiceNow Acquires Data.World to Enhance AI Data Governance Capabilities
ServiceNow has strategically expanded its AI and data management portfolio by acquiring data.world, a leading data catalog and collaboration platform, just two months after its acquisition of Moveworks, an AI-powered workplace automation company. This dual acquisition spree underscores ServiceNow’s commitment to embedding AI deeply into its workflow automation and enterprise data ecosystems.
The acquisition of data.world enhances ServiceNow’s ability to provide comprehensive data governance and discovery tools, enabling enterprises to unlock insights from disparate data sources while maintaining compliance and security. Combined with Moveworks’ AI-driven conversational automation, ServiceNow is positioning itself as a powerhouse for intelligent enterprise workflows. These moves come as ServiceNow reported strong revenue growth in Q1 2025, driven by increased demand for AI-enabled digital transformation solutions across industries.
Apple Explores AI Search Engine Integration in Safari, Signaling Potential Shift from Google
During a U.S. Department of Justice antitrust trial against Google, Eddy Cue, Apple’s Senior Vice President of Services, revealed that Apple is actively exploring the integration of AI-powered search engines into its Safari browser. Potential partners include OpenAI, Perplexity AI, and Anthropic. This move comes amid a reported decline in Safari’s search usage for the first time in 22 years, attributed to the rising adoption of AI tools.
Apple currently maintains a lucrative agreement with Google, reportedly worth $20 billion annually, to keep Google as the default search engine on Safari. While Cue acknowledged the financial benefits of this partnership, he also noted the transformative potential of AI in the search market. The announcement had immediate market implications, with Alphabet Inc.’s stock dropping over 9% following the news.
Figma Launches AI-Powered Tools for Streamlined Web and App Design
Figma has unveiled new AI-driven features aimed at simplifying the creation of websites, app prototypes, and marketing assets. The newly introduced Figma Sites allows users to generate and publish websites directly within the platform, complete with customizable elements like transitions, animations, and responsive design. An integrated content management system (CMS) enables easy editing of blog posts and management of assets such as thumbnails and slugs.
Additionally, Figma Make offers an AI-assisted environment for ideation and prototyping. Users can input prompts to create collaborative web applications, with the ability to modify elements through AI suggestions or direct code edits. These tools position Figma as a competitive player against platforms like Canva, Adobe, WordPress, Wix, and Replit, catering to both designers and marketers seeking efficient, AI-enhanced design solutions.AI Week # 19 has set the stage for an even more explosive future—are you ready for what’s next? From billion-dollar moves to everyday tech upgrades, the world of Artificial Intelligence is evolving faster than ever. Stay ahead by following our weekly AI digests, where we break down the top AI news, innovations, and industry shifts you can’t afford to miss. Don’t miss AI Week # 20—subscribe now and be the first to know what’s shaping tomorrow.
