Snowflake執行長預測2026年大型科技公司對AI的掌控將減弱，並提出其他關鍵趨勢

Snowflake CEO: Big Tech’s grip on AI will loosen in 2026 — plus 6 more predictions that will define the year

Over the past year, AI has begun reshaping work in tangible ways, with coding assistants that speed software development and chatbots that handle routine customer inquiries. But 2026 will be the year organizations move beyond these initial use cases to deploy systems that can reason, plan, and act autonomously across core operations.

This next stage has the potential to deliver dramatic gains, driven by shifts already underway in how AI models are built and deployed. The following predictions outline how the landscape will evolve in 2026 — from wider access to competitive models to new standards for measuring AI reliability — and how successful organizations will differentiate themselves to capitalize on these changes.

1 – Big Tech’s Grip on AI Models Will Loosen

For years, conventional wisdom held that only a handful of tech giants could afford to build competitive AI models. In 2026, that will change. New approaches to training like those developed by DeepSeek have shown that building the biggest, most expensive models isn’t the only path to strong performance. Companies are now taking open-source foundation models and customizing them with their own data, creating a faster, cheaper route to competitive AI. This democratization means far more organizations will create their own tailored models instead of relying solely on OpenAI, Google, or Anthropic.

2 – AI Will Have Its ‘HTTP’ Moment With a New Protocol for Agent Collaboration

Much as HTTP allows websites to connect freely across the internet, a dominant AI protocol will emerge next year that will allow agents to work together across different systems and platforms. This move towards standardization will unlock the true potential of agentic AI by allowing specialized agents from different providers to communicate and collaborate without vendor lock-in. Organizations will finally be able to build interconnected AI ecosystems rather than siloed applications tied to single providers. The age of the proprietary AI walled garden is ending.

3 – Teams That Resist ‘AI Slop’ Will Dominate the Creative Landscape

In 2026, a divide will emerge between those who use AI to amplify their own creativity and those who use it as a crutch. One group will leverage AI to expand their creativity and push their own ideas further and faster. The other will take the easy route, churning out generic content that floods the market but doesn’t resonate with customers. Organizations that take the former approach — empowering people to think strategically and use AI to enhance, rather than replace, their own creativity — will dominate their industries.

4 – The Best AI Products Will Learn From Every User Interaction

In 2026, the most successful AI products will build in continuous learning from user behavior. Much as Google’s search algorithm improved itself by learning which websites users actually clicked on, AI systems that capture feedback loops — like coding copilots do now when users accept or reject suggestions — will improve far faster than static models. Embedding these feedback loops into products will make increasingly complex use cases possible. Companies that take advantage of this continuous learning will gain compounding advantages.

5 – Enterprises Will Demand Quantified Reliability Before Scaling AI Agents

Business-critical AI applications require precise, measurable accuracy, not probabilistic answers. While consumer AI can afford to occasionally get things wrong, enterprise systems need exact answers to questions like “How much revenue did we generate yesterday?” In 2026, organizations will insist on systematic methods to measure the accuracy of agents before deploying them at scale, which will drive rapid innovation in sophisticated evaluation frameworks. Establishing these domain-specific testing standards will be essential for taking agentic AI from pilot projects to core business operations.

6 – Ideas, Not Execution, Will Become the AI Bottleneck

As AI agents handle more of the actual work of building and implementing projects, organizations will be limited by the quality of their ideas more than their ability to execute on them. This shift will be both liberating and daunting. It allows teams to rapidly prototype and deploy solutions that once took months, but success depends on asking the right questions and setting the right direction. In 2026, as execution becomes commoditized, strategic thinking and vision will separate high-performing organizations from the rest.

7 – Shadow AI Will Drive Enterprise Adoption from the Bottom Up

Employees who select their own free AI tools will remain the primary driver of enterprise AI adoption in 2026. Rather than waiting for IT departments to sanction approved products, workers are using ChatGPT, Claude, and other consumer AI tools for their daily work, forcing organizations to catch up with formal policies and infrastructure. Smart enterprises will recognize this grassroots adoption as a signal of what works and build their AI strategies around employee-proven use cases. The future of enterprise AI is being written by individual contributors, not by mandates from the top.

The Real AI Race Starts Now

The organizations that lead in 2026 won’t be those with the most AI pilots or the biggest technology budgets. They’ll be the ones that treat AI as a strategic discipline — building evaluation frameworks, establishing trust through verified accuracy, and empowering employees to use these systems effectively. The technology is ready. Enterprises must now deploy it responsibly at scale.

The opinions expressed in Fortune.com commentary pieces are solely the views of their authors and do not necessarily reflect the opinions and beliefs of Fortune.

Sridhar Ramaswamy is CEO of Snowflake, the AI Data Cloud company.

© 2025 Fortune Media IP Limited. All Rights Reserved. Use of this site constitutes acceptance of our Terms of Use and Privacy Policy | CA Notice at Collection and Privacy Notice | Do Not Sell/Share My Personal Information
FORTUNE is a trademark of Fortune Media IP Limited, registered in the U.S. and other countries. FORTUNE may receive compensation for some links to products and services on this website. Offers may be subject to change without notice.