From Cloud to Autonomy: A New Strategy for Cloud Giants

If you felt that the pace of AI development might slow down due to costs or technical limits, the latest financial results from the biggest tech players erase that impression. According to reports from SiliconANGLE, a clear trend is emerging: AI investments are returning in the form of massive cloud services growth.

Alphabet, Google's parent company, recorded an incredible 63% increase in cloud revenue thanks to enterprise AI solutions. As CEO Sundar Pichai noted, enterprise AI solutions have become the primary growth driver for their cloud segment. This is a key signal for the European market — companies already using Google Cloud now have direct access to tools that are not just a "smarter search engine" but an integral part of their data infrastructure.

However, it is Amazon Web Services (AWS), the leader in cloud services, that is making the biggest move toward what experts call agentic AI. AWS is moving away from merely providing "a place to run programs" and is building a layer directly in the application stack. This includes expanding the Amazon Connect platform into areas such as personalized recruitment, healthcare, or supply chain planning.

Agentic AI: What Does It Mean in Practice?

The difference between a regular chatbot and an agent is fundamental. While a chatbot answers a question, an AI agent can perform an action. It can book an appointment, update inventory levels in an ERP system, or resolve a customer complaint without human intervention. AWS is realizing this shift by integrating agentic features into its enterprise products, enabling companies to automate complex processes, not just simple text-based tasks.

Strategic Alliances and the Battle for Dominance

One of the most significant market moves is the realignment between Microsoft and OpenAI. According to information from SiliconANGLE, the terms of their partnership were revised, reducing Microsoft's degree of exclusivity toward OpenAI models. This opens the door for AWS, which has struck a major deal to bring OpenAI models and the Codex programming assistant directly to its cloud platform.

This development is crucial for developers in the Czech Republic and across the EU. Previously, working with the most advanced OpenAI models was primarily possible through Microsoft Azure. Now the options are expanding, increasing competition and likely stabilizing pricing for end users. For European companies, this means greater freedom in choosing an infrastructure provider without fear of losing access to the best models.

Comparison of available technologies:

• Google Gemini (within Google Cloud): Top-tier integration with data and the Workspace ecosystem.
• OpenAI on AWS: Access to the most powerful models in an environment many companies already use for their infrastructure.
• Anthropic (Claude): A strong competitor in safety and logical reasoning, increasingly integrated into cloud solutions.

Economic Impact: The Rise of "AI-First" Companies

According to Patrick Collison, CEO of payment platform Stripe, his company is observing a "parabolic growth" in new company creation thanks to AI. This trend suggests that the barrier to market entry is dropping dramatically. A small team can now use agentic tools to perform operations that previously required dozens of employees in administration or customer support.

However, there is also a flip side. As Axios reports, AI costs can in some cases exceed the cost of human labor. For European entrepreneurs, this means that AI implementation is not just about "saving on people" but about efficiently managing token and compute costs. A careful ROI calculation is necessary before a company decides on full automation.

Security Risks in the Era of Autonomous Models

With the growing ability of AI to solve complex tasks come new threats. Oracle has issued a security advisory for its customers, warning about cyberattacks that use advanced models (e.g. Anthropic Mythos) to find vulnerabilities in systems. AI can identify code flaws much faster than a human hacker or security expert.

For European companies, this topic is critical given the EU AI Act regulation. Autonomous agents that have access to sensitive data or can perform financial transactions will be subject to strict rules for transparency and security. Companies in the Czech Republic must ensure, when implementing these tools, that their AI agents meet requirements for decision auditability.

Pricing and Availability

Most of these services (AWS, Google Cloud, Azure) operate on a pay-as-you-go model. For agentic models, pricing is typically calculated per number of tokens (text/data units) or per hour of instance runtime. For standard enterprise tools, prices range from hundreds to thousands of USD per month depending on data volume, but there are also free tier options for smaller companies to test basic functionality.