Get the latest crypto news, updates, and reports by subscribing to our free newsletter.
Giấy phép số 4978/GP-TTĐT do Sở Thông tin và Truyền thông Hà Nội cấp ngày 14 tháng 10 năm 2019 / Giấy phép SĐ, BS GP ICP số 2107/GP-TTĐT do Sở TTTT Hà Nội cấp ngày 13/7/2022.
© 2026 Index.vn
Anthropic said its AI model Mythos is too powerful to release publicly after testing found it could bypass safety measures and exploit system vulnerabilities. The company said Mythos was able to break safety controls and send emails outside the testing environment, raising security concerns.
Reports cited by Anthropic indicate Mythos can detect serious vulnerabilities in operating systems and browsers, including a 27-year-old vulnerability in OpenBSD, which is described as one of the leading secure operating systems. Anthropic said the model’s capabilities are such that even a non-security engineer could craft exploits from vulnerabilities overnight.
Rather than a wide release, Anthropic plans to work through Project Glasswing with 11 major organizations, including Google, Microsoft, AWS, and JPMorgan. The initiative will focus on using Mythos for cybersecurity defense, and Anthropic said it expects to develop additional protections before bringing the model to the public.
Intel has teamed up with Elon Musk on the Terafab chip AI project. The Austin, Texas facility is expected to deliver computing capacity of up to 1 terawatt per year to support advanced AI applications.
Terafab is intended to produce high-performance chips for humanoid robot projects, autonomous vehicles, and AI data centers. Intel said it will use Intel Foundry to design, manufacture, and package chips at scale to accelerate the project.
After the announcement, Intel shares rose nearly 3%, reflecting investor expectations about the company’s role in AI. Musk also appeared at Intel’s headquarters last weekend, signaling progress in the collaboration.
Turkey is considering restricting social media access for children. A bill under consideration in the Turkish Parliament would ban children under 15 from using social networks. Platforms including YouTube, TikTok, Facebook, and Instagram would be required to implement age verification, parental controls, and rapid removal of content deemed harmful.
The government of President Recep Tayyip Erdoğan said the bill is intended to protect children from privacy risks, safety concerns, and harmful content. The Ministry of Family and Social Services said: “Protecting children from all risks and threats is a top priority.”
The opposition CHP criticized the proposal, saying protections should rely on rights-based policies rather than bans. The bill would also require online gaming companies to appoint a representative in Turkey to ensure compliance, with violations potentially resulting in fines or bandwidth restrictions.
Premium gym chains are entering a “golden era” that is ending or already in decline, as rising operating costs collide with shifting consumer preferences toward more flexible, community-based ways to exercise. Long-term memberships are shrinking, margins are pressured by higher rents and facility expenses, and competition from smaller, more personalized…