Moves from Adobe, Oracle, OpenAI and Google highlight how enterprise tech is doubling down on artificial intelligence (AI) in workflows, infrastructure expansion, and developer enablement across creative, business and cloud ecosystems.
Creative Professionals Get Agentic AI
Adobe announced Tuesday (Oct. 28) at its MAX 2025 event that Creative Cloud apps such as Photoshop, Lightroom, Premiere and Illustrator will now include an AI Assistant powered by agentic AI to handle repeatable tasks and surface personalized recommendations.
Adobe’s announcement also noted more than 100 new features, including image upscaling, generative editing and batch-image processing via its Firefly suite. The updates are part of what Adobe calls “collaborative intelligence,” where human input and AI capabilities operate side by side.
Since Firefly’s debut, the platform has generated billions of images and effects, signaling the appetite for automation in creative industries. The latest release expands Firefly’s integration into Express and Acrobat, connecting creative design with document workflows.
For enterprise strategists, the launch signals a broader shift as creative teams move from being tool users to AI collaborators, expanding productivity spending from software licenses to subscription-based workflows.
Business Apps Go Full-Tilt on Agents
Oracle’s “Apps 2025” update brings embedded AI agents and a new AI Agent Marketplace into its Fusion Cloud Applications and Industry Applications, with over 600 embedded agents now available. Oracle’s blog post emphasized that customers’ data will not be used to train underlying language models and that role-based access remains intact. Each agent can be customized by function, and enterprises can deploy them on Oracle Cloud Infrastructure or on-premise systems.
By integrating agents natively into its business applications, Oracle aims to simplify adoption for nontechnical teams while creating a new monetization layer through its marketplace. Customers, including Milwaukee Tool and BHE Renewables, are already integrating the agents across workflows, Oracle said.
The marketplace model also establishes a new monetization pathway similar to an app store. Instead of paying for compute time, enterprises can license domain-specific agents tied to measurable ROI benchmarks. For Oracle, it marks a shift from offering AI as a feature to embedding it as a core productivity driver across enterprise software.
Billion-Dollar Infrastructure Bets
In a major infrastructure push, OpenAI and Oracle announced plans to build the “Stargate” data-center campus in Port Washington, Wisconsin, as part of a broader $500 billion, 10-gigawatt compute initiative. Reuters reported that the Wisconsin campus is expected to deliver more than 4.5 gigawatts of capacity and create about 4,000 construction jobs.
At the same time, Anthropic expanded its partnership with Google Cloud to access over 1 million tensor-processing units beginning in 2026. Reuters covered the deal, noting it signals escalating demand for specialized AI hardware. Google said the agreement will help Anthropic train its Claude models faster and cut inference costs by about 40 percent.
These investments underscore how infrastructure spending has become the backbone of AI competitiveness. As training costs rise, hyperscalers and model providers are increasingly forming joint ventures to control power supply, chip sourcing and network distribution, effectively building the next generation of global compute grids.
Developers First
Google introduced “Vibe Coding” in Google AI Studio, enabling users to describe an app in natural language and have the system automatically configure models and APIs. Google’s blog post details the update along with a revamped App Gallery and annotation mode for visual customization. Vibe Coding reduces technical barriers for both developers and nondevelopers, allowing teams to prototype AI apps in hours instead of weeks.
The launch complements Google’s broader ecosystem, including Gemini Code Assist, Vertex AI and AppSheet, all aimed at accelerating the development of AI-native software. The launch also positions Google to compete more directly with Microsoft Copilot Studio and OpenAI’s custom GPT platform in the battle for developer adoption.
Source: https://www.pymnts.com/