AI Powers 30% of Microsoft Code—How Automation is Shaping the Industry

Artificial intelligence is rapidly reshaping the tech and business landscape, from writing a significant share of Microsoft’s code to fueling billion-dollar acquisition strategies and intensifying legal battles over data scraping. As industry leaders like Satya Nadella and Mark Zuckerberg reveal the accelerating integration of AI into core operations, publishers and policymakers scramble to adapt to new challenges and opportunities. This press review explores the latest developments in AI-driven software development, the escalating fight against AI bots, groundbreaking protocols for DevOps, major investment moves, and the evolving regulatory landscape for AI hardware exports.

AI Now Writes Up to 30% of Microsoft’s Code, Says Satya Nadella

Microsoft CEO Satya Nadella revealed that as much as 30% of the company’s code is now written by artificial intelligence. Speaking at Meta’s inaugural LlamaCon AI developer event in Menlo Park, California, Nadella stated, “I’d say maybe 20%, 30% of the code that is inside of our repos today and some of our projects are probably all written by software.” He emphasized that the proportion of code generated by AI at Microsoft is steadily increasing.

During the same event, Meta CEO Mark Zuckerberg shared that Meta is developing an AI model capable of building future versions of its Llama family of AI models. Zuckerberg estimated that within the next year, “maybe half the development is going to be done by AI, as opposed to people, and then that will just kind of increase from there.” The discussion highlighted a broader industry trend, with Google CEO Sundar Pichai previously stating that more than 25% of new code at Google is written by AI. Other companies, such as Shopify and Duolingo, are also shifting towards AI-driven workflows, requiring employees to demonstrate that AI cannot perform a job before new hires are approved.

Company	AI-Written Code (%)	Source
Microsoft	20-30%	CNBC
Google	25% (new code)	CNBC

Key Takeaway: AI is rapidly transforming software development at major tech companies, with up to 30% of Microsoft’s code and over 25% of Google’s new code now generated by AI, according to CNBC.

Publishers Struggle to Block AI Bots Amid Mounting Lawsuits

As legal battles over AI scraping intensify, publishers are finding it increasingly difficult to prevent AI companies from extracting their content. Ziff Davis is the latest media company to sue OpenAI for copyright infringement, highlighting the broader challenge: publishers lack reliable methods to stop AI bots from scraping their websites for free. Despite the use of robots.txt files, paywalls, and AI-blocking tags, many bots ignore these protocols or disguise their identities, making enforcement nearly impossible.

According to a recent Tollbit report, AI bot scrapes bypassing robots.txt increased by over 40% between Q3 and Q4 2024. For example, travel news site Skift, even after disallowing GPTBot via robots.txt, was still scraped about 60,000 times a week. Ziff Davis reported that GPTBot activity “significantly increased” even after requesting OpenAI to stop. The Wikimedia Foundation noted a 50% rise in infrastructure costs since January 2024 due to bots and AI scrapers. Lawsuits, such as those led by The New York Times and Ziff Davis, are helping define legal boundaries, but most publishers lack the resources to pursue similar action.

Metric	Value	Source
Increase in AI bot scrapes bypassing robots.txt (Q3 to Q4 2024)	40%+	Digiday
Skift weekly scrapes by GPTBot (after disallowing)	~60,000	Digiday
Wikimedia Foundation infrastructure cost rise (since Jan 2024)	50%	Digiday

Cloudflare reports over 800,000 sites have blocked all AI crawlers.
Paywalls are not fully effective; even paywalled sites like The New York Times and Skift remain vulnerable.

Key Takeaway: Despite new tools and legal action, publishers face escalating challenges in blocking AI bots, with scraping activity and associated costs rising sharply, as reported by Digiday.

MCP Protocol Set to Revolutionize DevOps AI Integration

The Model Context Protocol (MCP), introduced by Anthropic as an open standard in late 2024, is poised to transform how DevOps teams leverage AI. MCP enables AI models to connect directly to external tools and data sources, automating a wide range of actions beyond traditional code generation. This includes tasks such as modifying cloud service configurations, deploying applications, and managing databases through natural language prompts.

MCP operates on a server-client architecture, where MCP servers execute actions and MCP clients (AI agents) mediate between users and servers. This standardization allows developers to integrate AI into diverse workflows, automate connections to custom data sources, and enhance DevOps tools with contextual intelligence. For example, DevOps teams can use MCP to automate cloud management tasks, such as identifying and securing publicly accessible storage buckets, or to ingest and search documentation databases across platforms like Confluence or SharePoint.

MCP supports automation of complex DevOps tasks using natural language.
It enables model-agnostic connectors for documentation ingestion and workflow automation.
Security and usability challenges remain, such as the need for local server setup and risk of sensitive data exposure.

Security concerns include the potential for MCP agents to leak sensitive data or perform unintended actions. Mitigation strategies involve hosting models locally and applying least-privilege access controls. Despite setup challenges, MCP is expected to become as integral to DevOps as CI/CD technologies.

Key Takeaway: MCP offers DevOps teams a powerful, standardized way to automate and enhance workflows with AI, though practical deployment requires careful attention to security and setup, according to cio.com.

Thrive Capital Launches $1 Billion AI-Driven Acquisition Vehicle

Thrive Capital is raising approximately $1 billion for a new company, Thrive Holdings, designed to acquire and operate businesses that can benefit from artificial intelligence. According to The New York Times, Thrive Holdings will focus on developing and buying start-ups, particularly in traditional industries such as accounting, and will use cash flow from these businesses to fund further acquisitions and investments.

Thrive Holdings is structured as a permanent capital vehicle, allowing it to raise additional funds over time and hold stakes in companies indefinitely. The firm has already backed companies like Crete (accounting) and Long Lake (homeowner association management), and is investing in complementary sectors such as IT service providers. Thrive Capital, founded by Joshua Kushner, has a history of incubating successful companies, with at least five reaching valuations above $1 billion, including Oscar Health and Cedar.

Thrive Holdings’ initial funding round: about $1 billion.
Focus: Acquiring and operating companies in everyday industries, leveraging AI for efficiency.
Permanent capital structure allows for long-term ownership and ongoing fundraising.

Other venture firms, such as General Catalyst and 8VC, are pursuing similar AI-enabled roll-up strategies, but Thrive Holdings distinguishes itself by its operational focus and long-term investment horizon.

Key Takeaway: Thrive Capital is making a significant $1 billion bet on AI-driven acquisitions, targeting traditional industries and aiming for long-term operational involvement, as reported by The New York Times.

Trump Officials Consider Changes to Biden’s AI Chip Export Rule

According to Reuters, officials from the Trump administration are considering modifications to President Biden’s AI chip export rule, which is set to take effect on May 15. The current rule divides the world into three tiers, with most countries subject to export caps on advanced AI chips. Sources indicate that Trump officials are contemplating replacing the tiered system with government-to-government agreements to control global access to AI chips.

The potential changes reflect ongoing debates over how best to regulate the export of critical AI hardware, balancing national security concerns with the interests of U.S. technology companies. The rule’s implementation and any subsequent adjustments could have significant implications for the global AI supply chain and international relations.

Biden’s rule: Three-tier system for AI chip export controls, effective May 15.
Trump officials may shift to bilateral agreements for export regulation.

Key Takeaway: The U.S. government is actively reviewing its approach to AI chip export controls, with possible changes to Biden’s upcoming rule under consideration by Trump officials, according to Reuters.

Einschätzung der Redaktion

The rapid integration of AI into core software development processes at leading technology companies marks a fundamental shift in how digital products are built and maintained. When up to 30% of code at a major player like Microsoft is now AI-generated, it signals not only a dramatic increase in productivity but also a redefinition of the developer’s role. This trend is likely to accelerate, driving both efficiency gains and new challenges in code quality, security, and workforce dynamics. The growing reliance on AI for code generation will require organizations to rethink their talent strategies, invest in robust oversight mechanisms, and address potential risks related to intellectual property and system reliability. Ultimately, the competitive advantage will favor those who can best orchestrate the collaboration between human expertise and machine intelligence.

AI-generated code is rapidly becoming a core component of software development at scale.
Organizations must adapt to new workflows, oversight needs, and evolving talent requirements.
The balance between automation and human oversight will be critical for sustainable innovation.

Sources: