Anthropic may be using its specialization in coding and STEM as a differentiator in the buzzy frontier model market.

MMLU-Pro holds steady at 85.0, AIME 2025 slightly improves to 89.3, while GPQA-Diamond dips from 80.7 to 79.9. Coding and agent benchmarks tell a similar story, with Codeforces ratings rising from ...

Zimperium research finds many Android and iOS apps leak sensitive data, exposing enterprises to API attacks and hidden malware risks.

Although capable of reducing trivial mistakes, AI coding copilots leave enterprises at risk of increased insecure coding ...

Expert-managed API tuning delivers stronger security with less effort CAMBRIDGE, Mass., Sept. 24, 2025 /PRNewswire/ -- Akamai ...

A new report from Thales has found that API-based cyberattacks have spiked in the first half of 2025. The post 40,000+ API-Based Cyberattacks Identified in Massive 2025 Spike appeared first on Tech.co ...

Astra Security, the platform for continuous pentesting, is releasing its new API Security Platform, uncovering undocumented, zombie, and shadow APIs that put infrastructures and sensitive PII at risk.

API Security Testing Framework A powerful and stealthy API security testing tool designed to identify vulnerabilities in REST and GraphQL APIs with advanced evasion techniques and comprehensive ...

SwissBorg, a Switzerland-based crypto wealth management platform, said hackers exploited a vulnerability in the API of its staking partner Kiln, draining about 193,000 Solana tokens from its Earn ...

UltraViolet Cyber’s purchase of Black Duck’s application security testing business expands its unified security operations and highlights growing risks from AI-generated code, open-source reliance and ...