Anthropic Relaxes Safety Rules and Introduces AI Tools Targeting IBM

Anthropic, previously positioning itself as the most responsible player in the artificial intelligence market, has made a shift in its Responsible Scaling Policy. The safe scaling policy has been changed to allow continued work on potentially dangerous technologies if competitors possess analogous solutions. The company's board argues that shifting the focus to economic competitiveness at the federal level necessitates moving away from previous restrictions. Concurrently, the startup presented a new product offensive aimed at the enterprise sector. A key element is the Claude Code tool, which can analyze and modernize code written in the COBOL language. Anthropic claims that modernizing systems, which once required an "army of consultants" and years of work, can now be completed within a few quarters thanks to AI. This declaration directly targeted the business foundations of IBM, whose mainframe systems form the basis for bank transactions worldwide. The COBOL language, although created over 60 years ago, still handles about 95% of transactions at US ATMs. For decades, IBM maintained market dominance because the cost and risk of rewriting these systems into modern languages were an insurmountable barrier for most institutions.The market reacted violently. IBM shares lost over 13% of their value on Monday, translating to the evaporation of nearly $32 billion in market capitalization. The sell-off also affected the cybersecurity sector, including companies like CrowdStrike and Datadog, following Anthropic's announcement of code-scanning features for security vulnerabilities. Nevertheless, a slight rebound occurred on Tuesday, fueled by news of new Anthropic partnerships with companies such as FactSet and DocuSign, which investors interpreted as an opportunity for increased efficiency in the business services sector. „From the beginning, we've said the pace of AI and uncertainties in the field would require us to adapt.” — Dario Amodei

Perspektywy mediów: Emphasizes the disruptive impact of technology on stable consulting jobs and the risks associated with loosening AI safety constraints. Highlights the necessity of strengthening US competitiveness and the benefits of modernizing outdated infrastructure systems.