American startup Anthropic has unveiled a suite of new tools, Claude Cowork and the Remote Control feature, targeting investment banking and engineering automation. Simultaneously, the company has accused the Chinese firm DeepSeek of massive copyright violations. The situation is exacerbated by White House reports of Beijing's use of the latest Nvidia chips despite sanctions, forcing Anthropic to revise its security policy to maintain a competitive edge.

Anthropic's New Business Tools

The launch of Claude Cowork and Remote Control aims to revolutionize work in investment banking, human resources, and software engineering.

Accusations of Data Theft

Anthropic claims that Chinese firms DeepSeek and MiniMax trained their models on data from the Claude bot, violating copyrights.

Loosening of Safety Regulations

Competitive pressure forced Anthropic to add caveats to its AI safety policy, raising concerns about the ethics of development.

Stock Market Shock for IBM

The giant's stock fell by billions of dollars due to fears that AI will take over the market for modernizing systems based on the COBOL language.

The artificial intelligence market has entered a phase of brutal competition, where technical innovations intertwine with accusations of economic espionage. Anthropic has unveiled Claude Cowork – a set of plugins dedicated to the financial, HR, and engineering sectors – and Remote Control, a mobile version of the Claude Code tool. These innovations aim to transform the bot into an autonomous agent performing complex office tasks. These agents have sparked enthusiasm on Wall Street, causing a rise in the indices of software companies that have recovered from earlier fears of market cannibalization by AI. Product successes are accompanied, however, by a sharp conflict on the US-China line. Anthropic has officially accused Chinese firms DeepSeek and MiniMax of "pillage" of their language models by training their own systems on data generated by Claude. Pressure from Chinese competition, which according to US administration officials possesses the latest Nvidia Blackwell chips, has forced Anthropic into a controversial move. The company has softened its previous safety commitments (ASSP), introducing caveats allowing for faster model deployment in the face of an "existential AI arms race." Since the 1950s, when the concept of artificial intelligence was born, the race for dominance in this field has been viewed as key to global economic and military advantage, similar to the space race during the Cold War. Simultaneously, an unprecedented sell-off of IBM stock occurred in the financial market. The corporation lost about 40 billion dollars in market value after investors concluded that Anthropic's new tools for converting COBOL code could render IBM's traditional modernization services obsolete. Experts warn, however, against excessive optimism, emphasizing that automatic translation of old code is not equivalent to its full modernization and optimization for modern systems. „We are seeing systematic harvesting of our model outputs by Chinese entities, which threatens the very foundation of fair competition in the AI sector.” — CEO and co-founder of Anthropic 40 mld $ — IBM's market loss due to AI competition fears

Mentioned People

  • Dario Amodei — CEO and co-founder of Anthropic
  • Aneel Bhusri — Co-founder and executive chairman of Workday