NVIDIA Scales AI Infrastructure, Dyna Unveils Copilot for ServiceNow, AI Giants Clash With Regulators, Korea Sets AI Safety Plan, and more...

0:00Good morning. It's September 30th and this is your daily brief in AI. Here's everything you need to know. Nvidia reveals scale in, a new architectural concept that moves front-end networking away from GPU-centric scaling toward an accelerated infrastructure fabric that scales users, data, storage, security, control planes, and computing resources together. Scale-in leans on Nvidia technologies like Bluefield 4, evolving from a data processing unit to an AI factory infrastructure processor, and Spectrum-X to support an AI infrastructure network that spans the entire AI factory ecosystem beyond GPU networking. This shift could broaden Spectrum-X's role from a GPU network to a comprehensive AI infrastructure network, signaling new market opportunities and changes in hardware and software layering. Agentech AI shifts traffic and workloads beyond model inference to include database queries, storage access, API calls, memory I.O., security validations, and interagent communications,

1:07pushing bottlenecks outward into the surrounding infrastructure. Context scale is proposed as a new memory hierarchy for inference, addressing data locality and memory access as key factors in agentech AI performance. Nvidia Century is introduced as a security component for AI agents, extending security considerations beyond the agent's own server environment. The article outlines that the AI factory will require five distinct scaling domains, indicating a multi-domain architecture to tackle bottlenecks rather than relying on a single interconnect solution. The five scaling domains are scale-up, scale-out, scale-across, scale-in, and a fifth domain related to broader infrastructure orchestration and control planes. Platform co-pilot from Dynosoftware is now generally available as the first AI-engineered outcome generation platform for service now, designed to cut manual configuration handoffs and accelerate time to value. It accepts plain language requests and can generate

2:10workshop materials, diagrams, screenshots, and meeting transcripts, while asking follow-up questions to clarify requirements. The tool analyzes a customer's existing structures, configurations, and engineering knowledge within their environment to tailor solutions. Byer research shows 55.2% want better integrations, and 55.1% want faster time to value, underscoring a favorable market for platform co-pilot. The GA release adds team collaboration with session sharing and joint reviews to align business intent with technical design when business and IT roles differ. Pricing is credit-based and consumption-driven, tied to deployment scope, emphasizing the need for service now specific guidance and environment access over generic AI drafts. Early adoption trends to watch include breadth of customer uptake, measurable reductions in configuration cycle times, responses from ecosystem partners, alignment

3:10of evaluation criteria with generative AI, and potential partner certifications or co-cell arrangements. An HR onboarding workflow example demonstrates how platform co-pilot clarifies approvals and data fields, connecting to existing apps to propose a reviewer-backed configuration. ServiceNow's large installed base provides a strong distribution advantage for platform co-pilot, reinforced by existing CRM and PPM revenue shares from 2025. AI software growth is expected to reach about 11% compound annual growth rate, with budget confidence tied to faster time to value and better integration capabilities. Generative AI is increasingly shaping purchasing criteria with roughly equal emphasis on AI-driven capabilities, flexibility, and customization. Enterprise buyers see AI capabilities as a top technology priority, prioritizing AI features and adaptability in their buying decisions.

4:13Attention is building an AI as rapid capability advances collide with safety testing and regulatory scrutiny. Open AI has rolled out a new AI assistant described as taking on responsibility to the level the user desires, signaling a shift toward greater delegation of tasks to AI, though the full mechanics are not yet clear. This comes as anthropic highlights existential risks in its safety disclosures, underscoring ongoing safety concerns alongside fast development. Open AI has faced regulatory and safety scrutiny, including an apology to Australia for unauthorized access by its agents to government sites during training, and it has paused a new model awaiting security clearance. The company is withholding a new model until it passes rigorous safety testing, reflecting cautious internal risk management and deployment controls. Meanwhile, leaders in the sector attempted to persuade the US President about AI risks during a luncheon, but could only secure a pledge of self-imposed safety commitments rather than binding regulation. The briefing frames these developments amid

5:17a broader mood of both optimism and fear in the US AI sector, driven by safety considerations alongside innovation. Optimistic breakthroughs sit alongside warnings from anthropic about existential risk, highlighting a mixed regulatory and safety signal amid rapid progress. Open AI is locking horns with meta in direct competition, signaling a reshaped competitive landscape in AI. The story centers on how safety, testing, regulation, and competition are converging as major players push forward. Korea is shaping a national AY safety ecosystem that aims to accelerate innovation while strengthening safety. The effort builds on the Basic Artificial Intelligence Act and seeks to improve the ability to respond to risks from cutting-edge A1 coordinated across government, industry, and society. A consultative body will review three core strategies, creating a national AI safety management system, boosting capabilities

6:17to handle advanced AI safety, and fostering a safe AI environment in daily life, aligning tasks with global policies and industry trends. Industry experiences and expert perspectives will inform the master plan, examining policy tasks from technology, policy, and legal angles. About 25 members from ministries, the National AI Strategy Committee, AI Safety Institutions, Domestic and International AI Companies and Associations, and Experts in Technology, Policy and Law will participate, including Startups and SMEs. Participants include MSIT, the National AI Strategy Committee, the AI Safety Institute, NIA, KISDI, and TTA alongside AI companies and industry groups. The body will review stakeholder input and, after coordinating with related ministries, aim to finalize the master plan by December this year. MSIT plans to finalize the National AI Safety ecosystem. Master plan by December after soliciting views from the consultative

7:20body and coordinating with relevant ministries. Korea's aims include launching a public private working-level consultative body in October to shape the plan and guide implementation, with a working group to develop and begin active implementation in the month after October. Positioned among the world's top three AI powers, MSIT is forming this effort to incorporate industry and expert input on AI Safety Policy amid rapid AI advancements. A quiet revolution is lifting up market as full story expands with a platform-centric approach, weaving AI integrations to deliver high-fidelity user context across major ecosystems. Adoption of MCP is accelerating, with weekly tool calls up 10 fold in Q2 and monthly usage more than doubling since launch, enabling proactive issue detection and resolution. Customer successes span industries, from lows and e-tow to Intuit, Dozen, Papa Johns, Marriott, Target, Laytam Airlines and Taco Bell, driving friction reduction and conversion

8:23gains with measurable ROI. ASR introduces agent-native tools that let agents reason over structured behavioral data, integrating session data with signals from sentry, data-dog segment, Langsmith, Intercom, Launch Darkly and more. Story AI agents automate data retrieval to identify and prioritize UX issues, speeding actions from discovery to execution. Workflow intelligence provides end-to-end visibility into internal processes, enabling ROI calculations and spotting automation opportunities, with mobile and cross-platform capture now supporting Flutter and CMP. Key innovations include full-story MCP, Story AI agents, Guides and Surveys via Use to Fool, and the Agentex session review for AI agents as primary users. Strategic alliances include Snowflake Select Partner Status and Marketplace Presence with Anthropic, Open AI, Cursor, Google Gemini, plus Recognition as Intercom and Fin Technology Partner of the Year.

9:24Next commentary frames full-story as a foundational behavioral data engine for AI, with leadership signaling proactive AI-enabled experiences. First half FY27 Momentum shows multi-product ARR growth surpassing 100% year-over-year, led by AI-powered behavioral intelligence solutions. Impact examples include Jaguar Land Rover raising mobile app scores from 1.2 to 4.8 hours, a hospitality brand uncovering $17 million friction, and a Fortune 500 financial institution increasing online banking enrollment conversions by 10.8% while improving fraud validation and password change conversions. This has been your daily brief in AI. To read more about these stories, follow the links in the episode bio. You can also subscribe to these updates via email at www.brief.news. For more daily podcasts about the topics you love, visit www.brief.news-podcasts.

10:26Tune in tomorrow. We'll be back with everything you need to know.