TLDRs
- Salesforce plans $300M Anthropic token spending to accelerate AI-driven coding in 2026.
- AI tools are boosting engineering speed while reshaping developer responsibilities at Salesforce.
- Company adopts multi-model strategy using Anthropic, OpenAI, and in-house AI systems.
- AI increases production code output by 30%, shifting delays to testing stages.
The spending will largely support AI-assisted coding and software development workflows, marking one of the company’s most aggressive commitments to generative AI infrastructure so far.
Benioff made the remarks during an appearance on the All-In podcast, emphasizing that AI is not replacing engineers but instead accelerating how quickly they can build and deploy software. The investment highlights Salesforce’s growing reliance on external AI models to complement its internal development stack.
Coding Efficiency Takes Center Stage
According to Benioff, AI-powered coding tools are already transforming how Salesforce engineers operate. Rather than focusing on manual coding from scratch, developers are increasingly using AI systems to generate, refine, and debug code at scale.
The company has previously signaled a shift in workforce priorities. In early 2025, Salesforce stated it would not hire new software engineers for the year, instead focusing on expanding sales capacity with plans to add between 1,000 and 2,000 sales staff. At the same time, the firm committed to hiring around 1,000 graduates and interns, suggesting a rebalancing of technical roles rather than a reduction in talent intake.
Benioff has also indicated that AI already handles a significant portion of internal workloads, estimating that automation contributes between 30% and 50% of operational tasks inside the company.
Multi-Model AI Strategy Expands
Salesforce’s $300 million commitment to Anthropic is not occurring in isolation. The company has been actively investing in Anthropic since its Series C funding round and continued participation through later stages, including the Series G round in early 2026.
However, Salesforce is not limiting itself to a single AI provider. Its Agentforce 360 platform allows customers to choose between multiple large language models, including those from Anthropic and OpenAI. This multi-model approach is designed to give enterprise clients flexibility while maintaining a unified platform experience.
In parallel, Salesforce continues to develop its own AI systems such as CodeGen, which powers tools like Einstein for Developers and Einstein for Flow. Employees also reportedly use a mix of external AI tools, including Google Gemini and GitHub Copilot, reflecting a broader internal strategy of model diversity rather than dependence on a single provider.
Productivity Gains Reshape Engineering Work
Internal data from Salesforce suggests AI is already having measurable effects on software production. The company reported that the volume of code reaching production has increased by roughly 30%, driven largely by AI-assisted development tools.
However, this acceleration has also shifted bottlenecks. Instead of slow coding, delays are increasingly appearing in later stages such as testing, integration, and code review. As a result, senior engineers are spending more time on system architecture and design decisions, while AI handles early-stage coding assistance, bug detection, and optimization suggestions.
Industry-wide, AI coding tools are becoming a baseline requirement for competitiveness in enterprise software development. Tools like Anthropic’s Claude Code are increasingly seen as strategic assets that can influence enterprise deals and platform adoption before broader ecosystem rollout.
Salesforce also emphasizes security and trust in its AI strategy. By integrating partners like Anthropic within its trust boundary, the company ensures that customer data and AI interactions remain within its secure environment while still leveraging external model capabilities.
AI-Driven Transformation Ahead
Salesforce’s planned $300 million Anthropic investment underscores a broader shift in enterprise software: AI is no longer an optional productivity tool but a core infrastructure layer. As the company continues expanding its multi-model strategy, the balance between human engineering and machine-assisted development is expected to evolve further, reshaping how enterprise software is built and maintained.
Get 3 Free Stock Ebooks
Discover top-performing stocks in AI, Crypto, and Technology with expert analysis.
- Top 10 AI Stocks - Leading AI companies
- Top 10 Crypto Stocks - Blockchain leaders
- Top 10 Tech Stocks - Tech giants
