IT News Roundup: Meta Cloud Expansion, Apple vs OpenAI, and Major Breaches - July 12, 2026

This week in IT, the cloud infrastructure landscape is shifting as Meta prepares to compete directly with AWS, Azure, and Google Cloud. The legal battles over AI intensify with Apple filing suit against OpenAI, while major data breaches at Accenture and KDDI underscore the ongoing challenges of enterprise cybersecurity. Google's latest infrastructure report reveals that most organizations are not yet ready for agentic AI at scale.

Meta Plans Cloud Business to Sell AI Computing Power

Meta Platforms is reportedly developing a cloud infrastructure business that would sell access to AI computing power and models to external customers, according to a Bloomberg report from July 1. The move would position Meta as a direct competitor to Amazon Web Services, Microsoft Azure, and Google Cloud — the three providers that have dominated enterprise cloud infrastructure for over a decade.

Meta has been building massive AI data center capacity in recent years to train its own large language models. The company reportedly sees an opportunity to monetize excess compute capacity by offering it to other AI developers and enterprises. Industry analysts note that Meta's aggressive approach to infrastructure investment could reshape cloud pricing and competition, particularly in the AI inference and training segments where demand continues to outpace supply.

Source: Bloomberg

Apple Sues OpenAI Over Alleged Trade Secret Theft

Apple has filed a lawsuit against OpenAI alleging trade secret theft, according to Cybernews reporting on July 11. The legal action marks one of the most significant confrontations between a major hardware company and an AI developer, joining a growing list of disputes over intellectual property in the AI industry.

The lawsuit comes amid a broader wave of legal challenges against AI companies from content creators, publishers, and technology firms. Apple's case reportedly centers on allegations that proprietary technology and internal systems were improperly accessed or replicated. The outcome could set important precedents for how AI training data and model development are regulated.

Source: Cybernews

Accenture Confirms Breach After 35GB of Source Code Stolen

Accenture has confirmed it suffered a cyberattack after threat actors began selling an archive of approximately 35GB of data allegedly sourced from the company. The breach was reported on July 9 by Cyber Security Review.

The stolen archive reportedly contains source code and other sensitive internal data. Accenture stated it is aware of the incident and has remediated its source. The breach is notable because Accenture is one of the world's largest IT consulting firms, and the exposure of its source code could have implications for the many enterprise clients whose systems were developed or maintained using those internal tools and frameworks.

Source: Cyber Security Review

KDDI Data Breach Affects Over 12 Million Users in Japan

Japanese telecommunications giant KDDI disclosed that a data breach exposed the email addresses and passwords of more than 12 million people. The attack targeted an email system used not only by KDDI but also by five other internet service providers operating under its infrastructure: STNet, JCom Co, Chubu Telecommunications, NIFTY, and BIGLOBE.

The breach highlights the risks of shared infrastructure in the telecommunications sector. A single compromised email platform can cascade across multiple service providers, amplifying the scope and impact of the attack. Users of the affected ISPs have been advised to change their passwords and enable multi-factor authentication where available.

Source: Privacy Guides

Anthropic Explores Custom AI Chip Partnership with Samsung

Anthropic has begun preliminary discussions with Samsung Electronics to manufacture a custom AI accelerator chip, potentially leveraging Samsung's 2nm process and advanced packaging technologies. The Information reported the exploratory talks on July 2, noting that Anthropic has hired specialized silicon engineers and is still defining chip specifications, power requirements, and integration into its server clusters.

The move follows a trend of AI companies pursuing custom silicon to reduce their dependence on NVIDIA GPUs and lower inference costs. If realized, a custom Anthropic-Samsung chip could offer significant performance-per-watt advantages for Claude model inference and training workloads, while also diversifying the AI semiconductor supply chain beyond TSMC.

Source: Tech Startups

Google Report: 80% of Organizations Unprepared for Agentic AI Infrastructure

Google's 2026 State of AI Infrastructure report, published on July 10, found that more than four in five organizations need to upgrade their technology stacks to support AI agents at scale. The report highlights that agentic AI — autonomous systems that can plan, execute, and iterate on tasks — places significantly higher demands on infrastructure than traditional AI workloads.

Key findings include the need for better orchestration platforms, improved observability tools, and more robust security controls around AI agent permissions. Organizations that delay modernization risk falling behind as competitors leverage agentic AI for operational efficiency gains. The report is particularly relevant for IT leaders planning their infrastructure roadmaps for the coming year.

Source: CIO Dive

Open Source Vulnerabilities Double as AI Adoption Accelerates

Black Duck's 2026 Open Source Security Risk Analysis (OSSRA) report reveals that open source vulnerabilities have doubled to an average of 581 per codebase. The report found that 87% of codebases are now at risk, with 65% having been hit by attacks. The surge correlates directly with the explosive growth in AI development, which heavily relies on open source libraries and frameworks.

Sonatype Research also reported reaching 1.8 million malicious packages logged by the end of Q2 2026, with npm accounting for 96.6% of malicious package counts. The findings underscore the growing challenge of supply chain security in the AI era, where developers pull in hundreds of dependencies that may contain undetected vulnerabilities or malicious code.

Source: Black Duck