Anthropic Pushes Into South Korea While Its Flagship Models Stay Grounded
Anthropic announced the opening of a new office in Seoul on June 18, alongside a series of partnerships with South Korean government agencies, universities, and major tech firms. The company signed a memorandum of understanding with South Korea's Ministry of Science and ICT to collaborate on public-sector AI adoption, model safety testing, and addressing AI-related cybersecurity threats.
Key partnerships include collaborations with SK Telecom, LG CNS, and Naver Cloud, as well as work with Korea's National AI Research Lab — a consortium that includes top universities such as KAIST, Korea University, Yonsei University, and POSTECH. South Korea has emerged as one of Claude's most active markets globally, with usage concentrated in technical and creative work.
The Seoul office is led by KiYoung Choi, who brings three decades of experience leading technology businesses across Korea. The expansion comes ahead of Anthropic's anticipated IPO.
The timing is notable: Anthropic's most powerful models — Fable 5 and Mythos 5 — remain offline after the U.S. Commerce Department issued an export control directive on June 12 ordering the company to restrict foreign nationals' access. Rather than implementing selective restrictions, Anthropic chose to suspend the models entirely. Fable 5 had been publicly available for only three days before the ban took effect — one of the fastest model-to-ban timelines in the industry's history.
At a Seoul press conference, Anthropic's Managing Director of International Chris Ciauri stated: "We are very confident that in the coming days, the models will become available again" — the most specific confidence signal Anthropic has given about a timeline for restoring access.
Sources: Anthropic, UPI, TechCrunch
150+ Cybersecurity Experts Demand Reversal of Fable and Mythos Ban
The backlash against the U.S. government's decision to restrict Anthropic's most powerful models is intensifying. An open letter organized by former Meta chief security officer Alex Stamos has now gathered more than 150 signatures from cybersecurity executives and technical leaders, addressed to Commerce Secretary Howard Lutnick and National Cyber Director Sean Cairncross.
The letter's central argument: the ban takes the best defensive tools away from cybersecurity professionals while adversaries continue to advance. "To pull the best capabilities away from defenders without a good reason when our adversaries are rapidly advancing is dangerous," the letter states.
Notable signatories include Casey Ellis (founder of Bugcrowd), Jon Callas (former Apple security design and architecture manager), and Paul Vixie (computer scientist and DNS pioneer). The letter reflects a growing consensus in the security community that broad model restrictions may do more harm than good by disarming defenders while doing little to limit adversarial access to comparable capabilities from other providers.
Sources: TechCrunch, Fortune
OpenAI's GPT-5.6 Expected to Launch This Month
OpenAI is gearing up to release GPT-5.6, the next iteration in its model lineup, with a launch expected before the end of June 2026. OpenAI's chief scientist has called it a "meaningful leap," particularly in advanced reasoning and agentic workflows.
Based on developer reports and early access signals, GPT-5.6 is expected to feature a context window of up to 1.5 million tokens — roughly a 43% increase over GPT-5.5 — along with improved long-context reasoning, better multi-step planning, and enhanced error recovery in real-world agent tasks. The model is also expected to deliver better token efficiency, potentially lowering operational costs for developers.
The release comes as OpenAI continues its strategic pivot from chatbots to agents. The company recently launched Workspace Agents for business customers and has been sunsetting its "Pulse" proactive updates feature in favor of scheduled task execution. OpenAI is also preparing for its own IPO later this year, making GPT-5.6 a critical product moment.
Sources: Geeky Gadgets, Tech Times
Visa Embeds Payment Network into ChatGPT
Visa announced at its Payments Forum in San Francisco that it has embedded its payment network directly into ChatGPT, allowing AI agents to search for products, compare prices, and complete purchases on behalf of users. The integration lets users link their Visa cards to ChatGPT and authorize AI-initiated transactions at any merchant that accepts Visa.
The feature includes safety guardrails such as spending limits, required approval steps, and approved merchant lists to minimize fraud. This represents a significant step beyond OpenAI's earlier "Instant Checkout" feature, which was retired in March after struggling with merchant adoption.
The move signals that AI-powered commerce is quickly moving from concept to infrastructure. If successful, it could establish a template for how payment networks integrate with AI agents across the industry — and raises important questions about consumer protection, impulse purchasing, and the role of AI in financial decisions.
Sources: ABC News, Yahoo Finance
AI Chatbot News Usage Reaches 10% Globally
According to the Reuters Institute's Digital News Report 2026, 10% of people worldwide now use AI chatbots for news consumption every week — up from 7% a year ago. While still a minority, the growth rate is significant and suggests that AI-mediated news consumption is becoming a mainstream habit faster than many in the news industry expected.
Meanwhile, a Pew Research Center study found that only 16% of Americans expect AI to benefit society over the next 20 years, with roughly two-thirds saying AI development is moving too fast and 67% expressing doubt that the government will regulate it effectively. The gap between growing usage and persistent skepticism highlights the tension at the heart of AI adoption: people are using these tools more even as trust in them remains low.
Sources: Build Fast with AI