The CFO AI Digest: October 14

Every big AI player is fighting for the enterprise.

This week’s Digest compares Google, Amazon, Adobe and Salesforce’s moves toward capturing more of the enterprise market with AI. Plus, OpenAI’s own AI chips and Figure 03’s most advanced humanoid yet.

Here’s the rundown:

1. Google launches Gemini platform for enterprises

Google announced Gemini Enterprise, a full-stack platform that embeds AI across enterprise workflows. It includes a no-code workbench for building AI agents, prebuilt assistants for research and analysis, and integrations with Salesforce, SAP, and Microsoft 365, in addition to Google Workspace.

CFO takeaway: Google just reminded us that the enterprise AI race is largely a distribution and reach game. By embedding Gemini across Google Workspace, with its 3B+ users, Google’s pitch to existing users is clear: minimizing change management and eliminating the need to purchase another tool translates to faster adoption and ROI. And with integrations beyond Workspace, it’s covering its bases elsewhere too. As a partner at a VC noted on Substack, “The battle for the enterprise is around who owns the workspace, and this is a part of Google’s rallying cry.”

2. Amazon releases agentic AI platform for enterprises

AWS introduced Amazon Quick Suite, an agentic AI platform that helps employees automate tasks, analyze data, and retrieve insights across internal tools and documents. It connects to popular systems like SharePoint, Salesforce, Google Drive, Adobe Analytics, and over 1,000 other apps via standardized integrations like OpenAPI and the Model Context Protocol (MCP). Employees can build custom agents without code, and use Quick Suite inside browsers or as a standalone app.

CFO takeaway: Quick Suite looks a lot like Google’s Gemini Enterprise, but Amazon is leveraging its proximity to enterprise data here. AWS hosts a large share of enterprise infrastructure and storage. If you’re an AWS customer, deploying Amazon’s AI means one consolidated vendor bill. If you’re not, this is Amazon’s move to capture more of your cloud spend. Amazon’s angle: your AI provider should be the one hosting your data.

3. Adobe unveils AI agents for B2B marketing and sales

Adobe launched a suite of AI agents designed to support B2B marketing and sales teams with targeting, personalization, and customer journey orchestration. The agents can identify buying group members, build campaigns, and surface insights from cross-channel data. Additional agents for account qualification and product discovery are coming soon, including a multimodal AI interface that supports voice, text, and image-based interactions.

CFO takeaway: In its enterprise AI push, Adobe is betting on marketing automation. It already powers campaigns for many companies through Marketo and Adobe Experience Cloud. Now, its AI agents can help B2B teams move faster and hand off better-qualified leads to sales. This could help shorten deal cycles and lower customer acquisition costs, two major levers for any sales-driven business.

4. Salesforce expands platform for enterprise AI agents

Salesforce announced the global rollout of Agentforce 360, a full platform for deploying AI agents across sales, service, marketing, IT, and more. It combines AI reasoning, automation, and conversational interfaces with data from Salesforce’s Customer 360, Slack, and partner apps. The platform supports voice, chat, and workflow agents, and integrates with tools like Google Drive, Snowflake, and OpenAI. With 12,000 customers live, Agentforce is designed to turn Salesforce into an AI-powered operating system for enterprise work.

CFO takeaway: Salesforce has focused on the front office stack, like CRM (Sales Cloud), customer support (Service Cloud), and employee collaboration (Slack). Now it’s embedding AI agents into those systems so they can act on existing data and workflows. That gives them the ability to respond to support tickets, generate marketing campaigns, schedule meetings, and more. The financial play: consolidate spend and automate more of the business through software in your existing stack.

5. OpenAI and Broadcom to deploy 10GW of custom AI accelerators

OpenAI announced a multi-year partnership with Broadcom to co-develop and deploy 10 gigawatts of custom AI accelerators, chips optimized for running and training large AI models, and networking systems. OpenAI will design the chips and systems, embedding its learnings from building frontier models. Broadcom will provide the ethernet-based infrastructure to scale them across OpenAI’s data centers and partner facilities.

CFO takeaway: This is OpenAI's second major infrastructure deal in two weeks, as it continues to build compute to keep up with AI demand. This time however, OpenAI is designing its own chips. While building custom silicon requires a large upfront investment, it will be cheaper than buying Nvidia's at scale. But if models evolve faster than the chip roadmap, OpenAI risks locking itself into hardware that no longer fits. Still, as one technical staff member noted, they're nine months from what could be the fastest production ramp-up for a first-time chip in the industry.

6. AI robotics company Figure launches new humanoid robot

Figure introduced Figure 03, its third-generation humanoid robot built to bring the company’s Helix vision-language-action AI system into the real world. The robot has a redesigned sensory and hand system, which enables better object manipulation and autonomous learning in complex environments. It has been engineered for high-volume manufacturing and will be produced at BotQ, Figure’s new manufacturing facility in California that’s expected to produce up to 100,000 units over four years.

CFO takeaway: We’ve seen AI scale digital work. Figure is applying that logic to the physical world. Scaling hardware is harder and means solving for reliability, supply chains, and real-world variability. The payoff, though, could be transformative: businesses shift labor costs to capital investments, while consumers replace many recurring service expenses with a one-time purchase.

See you next week.

Don't miss these

3 ways forward-thinking finance teams use data to drive strategy

At OnRamp New York, finance leaders from General Catalyst, Barry’s, and Crumbl highlighted the finance function's valuable asset: clean, comprehensive data.

The CFO AI Digest: October 8

OpenAI makes its biggest enterprise push yet, IBM and Deloitte deepen bets on Claude, AMD and OpenAI enter multibillion-dollar GPU deal.

How we’re scaling a lean finance team with AI — and what’s next

Brian Hogeland, CFO at national distributor Packer Fastener, shares how he embraces the 10x mindset to find creative ways for his small team to do more with less.