Nature

A new platform from data training company Scale AI will let artificial intelligence developers find their models’ weak spots.

The academic and author discusses what to expect from the singularity, the need for AI self-correcting mechanisms, and what hope there is for superintelligence safeguarding democracy.

The news follows the breakout success of DeepSeek and growing pressure from rivals like Meta.

Led by a former OpenAI executive, Amazon’s AI lab focuses on the decisionmaking capabilities of next-generation software agents—and borrows insights from physical robots.

Subscribers can now organize their conversations with the AI assistant into neat little folders.

An unsecured database used by a generative AI app revealed prompts and tens of thousands of explicit images—some of which are likely illegal. The company deleted its websites after WIRED reached out.

Elon Musk said he borrowed the name from a 1960s science fiction novel, but another AI startup applied to trademark it before xAI launched its chatbot.

Social platform X struggled after Elon Musk took over, but its fortunes improved dramatically after US President Donald Trump won reelection. Now it will be become part of Musk’s AI startup xAI.

Social Security systems contain tens of millions of lines of code written in COBOL, an archaic programming language. Safely rewriting that code would take years—DOGE wants it done in months.

Researchers looked inside the chatbot’s “brain.” The results were surprisingly chilling.