South Korea has secured access to Anthropic’s highly anticipated cybersecurity-focused artificial intelligence model, Mythos, marking a significant step in the country’s efforts to strengthen its digital defenses amid growing concerns that AI-powered cyber threats could become more sophisticated and harder to contain.

The development places South Korea among a select group of nations participating in Anthropic’s Project Glasswing initiative, an international program designed to deploy frontier AI systems for identifying, assessing, and helping remediate cybersecurity vulnerabilities before they can be exploited by malicious actors.

South Korea’s Ministry of Science and ICT made the announcement on Wednesday, saying that the government-backed Korea Internet & Security Agency (KISA) had successfully secured access to Mythos through its participation in the project.

The ministry said it had been working closely with Anthropic and confirmed KISA’s involvement in the initiative, which seeks to harness advanced AI models to detect software weaknesses at a scale and speed that would be difficult for human analysts to match.

The announcement follows a report by the Financial Times that Anthropic plans to broaden access to Mythos to approximately 150 organizations across more than 15 countries. According to the report, some of South Korea’s largest technology companies are included in the expansion, including Samsung Electronics, SK Hynix, and SK Telecom.

The move follows the rapid emergence of cybersecurity as one of the most strategically important applications of advanced AI. While much of the public focus has centered on generative AI tools for content creation, software development, and productivity, governments and corporations are increasingly concerned about the technology’s ability to uncover software flaws, network weaknesses, and security vulnerabilities at unprecedented speed.

That capability presents both opportunities and risks. Security teams can use AI to strengthen defenses and identify weaknesses before attackers do. At the same time, the same technologies could potentially enable cybercriminals and hostile state actors to discover and exploit vulnerabilities far more efficiently than in the past.

The issue has become particularly relevant for South Korea, which hosts some of the world’s most important semiconductor, electronics, and telecommunications companies. As global competition over advanced technologies intensifies, these firms have become increasingly attractive targets for cyber espionage, intellectual property theft, and state-backed hacking campaigns.

The country’s semiconductor sector alone occupies a critical position in global supply chains. Samsung Electronics and SK Hynix are among the world’s largest memory chip producers, supplying components essential to artificial intelligence infrastructure, data centers, smartphones, and advanced computing systems.

Access to Mythos could therefore provide South Korean institutions with a powerful new tool to identify vulnerabilities across critical digital infrastructure, industrial systems, and enterprise networks.

The ministry said South Korea would continue pursuing efforts to strengthen national cybersecurity capabilities by leveraging frontier AI technologies while also investing in domestic AI-driven security solutions.

Governments, as a strategy, are seeking to avoid dependence on foreign technology while still benefiting from the most advanced AI systems available globally. Policymakers view cybersecurity capabilities as a matter of national security, particularly as cyber threats become more automated and sophisticated.

The announcement also comes amid growing international debate over access to advanced AI models. European banks, technology firms, and government agencies have recently expressed concerns that restrictions on frontier AI systems could create competitive disadvantages in cybersecurity preparedness.

By securing participation in Project Glasswing, South Korea gains early access to technology that many organizations worldwide are still seeking to obtain. That access could help the country strengthen cyber resilience while giving local companies and researchers valuable experience working with some of the most advanced AI security tools currently available.

Following the release of Anthropic’s Mythos, AI has increasingly become a central component of both cyber defense and cyber offense. Thus, initiatives such as Project Glasswing are being viewed not simply as technology programs, but as strategic assets in a global competition to secure critical digital infrastructure.

Microsoft has sharpened the timeline in the global quantum computing race, unveiling a new quantum chip developed with the aid of artificial intelligence and declaring that commercially useful quantum computers could arrive by 2029.

The announcement places Microsoft alongside rivals racing to turn quantum computing from a scientific experiment into a commercially viable technology. The company now expects practical quantum machines to emerge within three years, matching a timetable recently outlined by IBM and intensifying competition with Alphabet, Amazon, and a growing number of Chinese quantum research initiatives.

At stake is a technology that could fundamentally alter computing. Quantum systems are expected to solve problems involving molecular simulations, drug discovery, advanced materials, logistics optimization, and cryptography at speeds unattainable by conventional computers. Governments and corporations view quantum computing as a strategic technology with implications for economic competitiveness and national security.

The centerpiece of Microsoft’s latest push is Majorana 2, a successor to the company’s first Majorana chip introduced last year. Unlike many competitors that rely on aluminum-based superconducting materials, Microsoft’s new chip uses lead, a material the company says delivers dramatic performance improvements.

According to Microsoft executive vice president Jason Zander, AI-powered materials science tools played a critical role in identifying and engineering the new approach. The company says the redesign generated a thousand-fold improvement in certain performance metrics compared with the previous generation.

“The reason why people don’t use it to build chips is it requires an incredibly specialized process to be able to go figure that out. And we figured it out,” Zander said.

Currently, AI is increasingly being used not only to develop software but also to accelerate scientific discovery and engineering. Technology companies are investing heavily in AI-driven materials research, hoping to shorten development cycles for semiconductors, batteries, pharmaceuticals, and next-generation computing systems.

For Microsoft, the development also strengthens a broader strategy that combines leadership in artificial intelligence with ambitions in quantum computing. If successful, the company could eventually integrate quantum capabilities with its cloud and AI platforms, creating a powerful competitive advantage in enterprise computing.

Yet Microsoft’s announcement also revives long-running scientific disputes surrounding its quantum research. The company’s quantum architecture relies on exotic quasiparticles known as Majoranas. Microsoft has long argued that these particles could enable more stable quantum computers that are less prone to errors than competing approaches.

However, the existence and practical implementation of Majorana-based quantum systems have remained subjects of intense debate within the scientific community. Microsoft previously faced scrutiny over research related to Majorana particles, and some physicists continue to question whether the company has provided sufficient evidence to support its claims.

Henry Legg, a lecturer in quantum physics at the University of St. Andrews, challenged the company’s latest announcement.

“Microsoft can use as much lead as they like – it is not going to shield them from the basic scientific principle that your results need to be reproducible,” he said.

Critics argue that independent researchers need access to more experimental data to verify Microsoft’s findings. Some have pointed to concerns surrounding earlier studies and say similar questions remain unresolved.

Microsoft counters that much of the underlying data cannot be publicly released because of intellectual property concerns. Company executives say detailed information has been shared with the U.S. Defense Advanced Research Projects Agency (DARPA), which is evaluating multiple quantum computing approaches.

“We’ve done enough of the physics to really have great data,” Zander said. “Believe me, I would not spend the money on the engineering if I felt like we were still off on the physics.”

As billions of dollars flow into quantum research, companies face pressure not only to demonstrate scientific progress but also to convince investors, customers, and governments that practical systems are within reach.

Just weeks ago, IBM announced plans to invest $10 billion in quantum computing and outlined its own roadmap toward fault-tolerant systems. Meanwhile, China continues to pour resources into quantum research as part of its drive to achieve technological self-sufficiency and reduce dependence on Western technologies.

OpenAI rival Perplexity CEO Aravind Srinivas has zeroed in on what he believes will ultimately determine the winners in the artificial intelligence race: the ability to deliver maximum economic value from the energy consumed by AI systems.

In a CNBC interview on Wednesday, Srinivas argued that the companies best able to maximize “token value per watt per user”, essentially delivering the highest useful output per unit of energy and per user, will command the highest valuations in the long term.

A token is the basic unit of data that an AI model processes when handling a query or task. Each token requires computational power and, by extension, electricity.

“Whoever is able to maximize this particular objective really will, by balancing accuracy, latency, cost, privacy and intelligence all together, they’re going to win. That’s what’s going to win long term,” Srinivas told CNBC’s Elaine Yu.

He acknowledged that current AI spending patterns raise legitimate questions. Many companies are pouring billions into infrastructure with limited visibility into returns, creating significant waste.

“You hear companies saying, I am spending a ton of money on AI. And I know some great stuff is happening, but I know there’s a ton of waste. How long do I have to wait for it to really show up in revenue, and how long do I have to wait to really get the costs under control?” he asked.

Srinivas views this as a temporary phase. He believes the industry will quickly improve efficiency, but the companies that solve the energy-to-value equation first will build durable advantages.

The Shift Toward Agentic AI and Orchestration

Perplexity is positioning itself at the center of this efficiency drive through a strong focus on agentic AI — systems capable of handling complex, multi-step tasks autonomously rather than simple prompt-and-response interactions. In February, the company launched Perplexity Computer, an agent designed to execute extended workflows.

A key innovation is the newly introduced Personal Computer tool, which Perplexity calls an “orchestrator.” This system intelligently decides which AI model to use for a given task, how different agents should collaborate, and whether processing should occur locally on a device or in the cloud.

On Wednesday, Perplexity announced that Personal Computer is now available on Microsoft’s Windows operating system, allowing integration with apps like Word and Outlook, in addition to its existing availability on Apple’s Mac platform.

Srinivas emphasized the importance of this layer, saying: “The data center is coming to your laptop… this is an orchestration problem. We believe that by solving that, we’ll be building a pretty valuable company that has endurable, long-term advantage.”

By acting as a neutral orchestrator that works across different models (including those from Anthropic), chips, operating systems, and hardware providers, Perplexity aims to create a versatile “AI operating system” that optimizes for multiple objectives simultaneously — cost, speed, privacy, and performance.

Platform-Agnostic Strategy in a Crowded Field

Perplexity, last valued at around $20 billion, trails far behind larger rivals such as Anthropic (nearing $1 trillion) and OpenAI (over $850 billion). Anthropic confidentially filed for an IPO this week, highlighting intense investor appetite for AI companies.

The competitive landscape is intensifying. OpenAI, Anthropic, and Google are all ramping up agentic capabilities, while Microsoft and Apple are building their own AI agents and assistants. Microsoft unveiled new coding and reasoning models on Tuesday, and Apple is updating Siri using Google’s AI technology.

Srinivas remains confident in Perplexity’s differentiated approach.

“I think they absolutely will try to build their own AI systems, but we believe we’re building the most versatile operating system by making it work across different models, across different chips, across different traditional operating systems, different hardware providers, different laptops. That hybrid neutral orchestration layer is what we are doing, and that allows us to balance all the different objectives simultaneously,” he said.

He noted that Perplexity has tripled its annualized revenue since the beginning of the year, largely “thanks to model advances that have been made by Anthropic,” whose models are integrated into Perplexity’s platform. This highlights a key strength of the orchestration strategy: the company benefits from rapid improvements by frontier model providers without bearing the full cost of developing them in-house.

Why Efficiency Will Define the Next Phase of AI

Srinivas’s emphasis on energy efficiency and orchestration addresses a growing tension in the AI industry. Hyperscalers are spending hundreds of billions on data centers and chips, yet real-world utilization rates remain low in many cases, and returns on investment are still being proven at scale. The focus on edge computing, running AI locally on devices like laptops and phones, could dramatically reduce power consumption, improve speed and privacy, and lower latency.

This shift from cloud-only to hybrid cloud-edge architectures represents a potential inflection point. Companies that master orchestration are considered to be better positioned to deliver cost-effective, responsive AI experiences while minimizing environmental impact and energy costs — factors increasingly important to both regulators and corporate buyers.

For Perplexity, this strategy aims to create a sustainable moat in a field where raw model performance alone may not guarantee long-term leadership. The company hopes to carve out a valuable niche as the “AI operating system” layer by focusing on intelligent routing and system-level optimization rather than solely chasing the largest models.