The week’s Big Tech developments point to a maturing artificial intelligence (AI) ecosystem. The focus is shifting from building smarter models to building the pipelines, agents and data platforms. These elements make models work together at scale.

Salesforce Goes All In on the Agentic Enterprise

At its flagship conference in San Francisco, Salesforce made its most aggressive bet on where enterprise AI is headed. The company unveiled Agentforce 360, its new platform for what CEO Marc Benioff calls the “agentic enterprise.” It is designed to connect humans, agents and enterprise data on one trusted system. According to Salesforce’s official announcement, Agentforce 360 represents “the evolution of CRM.” It embeds autonomous and multimodal AI across its Service, Marketing and Commerce clouds.

The launch builds on expanded partnerships with OpenAI and Anthropic to bring their frontier models, including GPT-5 and Claude, into Salesforce’s platform. This addition powers intelligent workflows for sales, finance and customer service. These models will enable domain-specific AI agents capable of handling tasks such as forecasting, contract analysis and lead management. They do so while adhering to enterprise compliance requirements.

Salesforce also backed its AI expansion with a $15 billion investment in San Francisco over the next five years. This investment aims to grow its innovation footprint and workforce. It reaffirmed its long-term goal of $60 billion in annual revenue by 2030. The company highlights that agentic workflows and automation will drive the next wave of enterprise growth.

Dreamforce announcements also introduced Slack’s expanded role as the interface for Salesforce’s agent ecosystem. Additionally, there are new Einstein Copilot features and updates to its Trusted AI governance framework. These updates focus on data provenance, auditability and model explainability. These moves cement Salesforce’s strategy to make AI integral to how every business unit collaborates. They also guide how decisions are made.

Adobe Turns Generative AI Into a Branded Service

While Salesforce reimagines workflows, Adobe is reinventing creative infrastructure for the AI era. The company launched Adobe AI Foundry. This platform lets enterprises train custom generative models on their proprietary brand assets, including video, 3D design and text. It ensures every AI-generated output matches their identity and tone.

Advertisement: Scroll to Continue

Early adopters, such as Home Depot and Walt Disney Imagineering, are already piloting brand-specific models through Foundry. They use this platform to automate campaign creation and content workflows at scale. The company is shifting toward usage-based pricing tied to generative output instead of traditional software licenses. This move is reported by Investors.com, as part of its broader AI-as-a-service transition.

Adobe says Foundry models are built on top of its Firefly foundation family. However, they are trained in a way that keeps enterprise IP secure and isolated from public data sets. This focus on provenance and content authenticity addresses one of the biggest risks in generative AI: brand dilution through synthetic media. The initiative positions Adobe not just as a creative software provider but as a brand infrastructure company for the AI economy.

Oracle is Rebuilding the Data Foundation for AI

As peers move up the stack, Oracle is reinforcing the layer beneath it. At the Oracle AI World event, the company introduced the Oracle AI Data Platform, Autonomous AI Lakehouse and Oracle Database 26ai. These are three pillars designed to merge data governance, analytics and AI in one environment.

The new architecture brings vector search and in-database agent frameworks directly into Oracle’s data systems. This allows enterprises to run generative and predictive AI workloads without moving sensitive data into external stores. Oracle’s pitch is clear: Bring AI to the data, not the other way around. The company’s newly unveiled Autonomous AI Lakehouse, meanwhile, is designed to unify structured and unstructured data with native AI governance capabilities. This is a response to enterprise demand for transparency and security across hybrid and multicloud environments.

To power these offerings, Oracle expanded its infrastructure partnership with AMD. This ensures access to the next generation of graphics processing units optimized for large AI workloads on Oracle Cloud Infrastructure. The collaboration allows customers to scale AI training and inference workloads across cloud and on-premises environments. Oracle executives said this approach enables faster model deployment, lower data latency and complete control of compliance pipelines. It targets enterprise CIOs, who must balance innovation with regulation.

Google Expands the AI Infrastructure

Completing the week’s major announcements, Google underscored that AI’s future depends as much on physical infrastructure as it does on algorithms. The company announced a $9 billion investment through 2027 to expand its AI and cloud footprint in South Carolina, part of a larger $24 billion global program spanning the United States and India. The investment will fund hyperscale data centers, subsea cables, renewable-energy capacity and new fiber networks to support the exponential computing needs of frontier AI models.

This expansion follows Google’s $15 billion investment in India announced in October, which includes an AI hub in Visakhapatnam. Both initiatives reflect Google’s belief that controlling the physical layer of computing, connectivity and energy is essential to maintaining leadership in AI services.