World on the Brink of Destruction, US Arms Supplier Reveals Horrifying Facts

Anthropic has recently become a frequent headline maker. The supplier of ‘advanced weaponry’ in the form of artificial intelligence (AI) tools for the US military has come under scrutiny after daring to oppose the United States government. As a result, Anthropic was ‘kicked out’ and placed on a supply chain risk list. The company has since drawn further attention following the launch of its latest model, named Mythos.

Many nations are immediately concerned by the sophistication of Mythos, which is said to be capable of rapidly identifying system vulnerabilities, potentially allowing for exploitation by cybercriminals. In a recent warning, Anthropic called on major AI companies to delay further AI development, warning of a potential global catastrophe. According to the company, AI technology is evolving too rapidly; if this continues, AI capabilities may increase faster than society’s ability to mitigate the resulting risks, as reported by Reuters.

The creators of Claude stated that AI’s ability to complete tasks autonomously has doubled approximately every four months and is heading towards ‘recursive self-improvement’—a point where technology can evolve without human intervention. ‘When [AI] systems are capable of fully building their successors, the way we secure, monitor, and shape their behaviour becomes far more critical,’ Anthlypic stated in a blog post.

The company believes that delays in AI development would allow society to be better prepared for future implications. ‘We are not there yet, and recursive self-improvement of AI is not a certainty. However, it could happen faster than most institutions expect,’ wrote Anthropic co-founder Jack Clark and Anthropic Institute leader Marina Favaro.

Concerns that advanced AI systems could escape human control and cause social damage have increased alongside technological sophistication. Anthropic’s Mythos model itself shook various industries, including banking and software, earlier this year with its ability to find vulnerabilities in existing code. Unfortunately, current regulations remain very slow, particularly in the US, the home of leading AI technology giants. An executive order from the Trump administration earlier this week placed responsibility on the companies themselves, requesting they voluntarily submit their most powerful models for government cybersecurity testing before public release.

AI researchers have previously urged a pause, though with less success. Elon Musk, owner of the AI laboratory xAI, was among those supporting the 2023 push by the non-profit Future of Life Institute to halt AI development for six months to allow for safety measures. Anthropic has long positioned itself as a safety-focused AI laboratory. Earlier this year, they refused to allow the US military to use their models for domestic surveillance and fully autonomous weapons, triggering a backlash from the government, which placed them on a national security blacklist effective by the end of 2026.

Nevertheless, Anthropic continues to release increasingly sophisticated models. In February, they rescinded a key safety pledge, stating they would no longer withhold potentially dangerous AI if competitors approached similar capabilities. Recently, the company was valued at US$965 billion in a massive funding round and has secretly filed for an initial public offering (IPO) in the US, placing it ahead of its competitor, OpenAI, in terms of both valuation and the race for critical funding.