Global AI Industry Does Not Meet Security Standards, Report Warns

Anthropic.
Reuters
The American artificial intelligence (AI) company Anthropic obtained the highest score in a biannual security ranking, but on a global level the sector is unable to combat "existential" threats, according to a report published this Tuesday (7).
The Future of Life Institute, an American AI security think tank, evaluated nine of the world's leading AI companies and defined a ranking based on public data and information provided by the companies themselves.
The nine companies are failing to combat "existential" threats, such as the development of models that achieve a human-like level of intelligence, known as "artificial general intelligence" or AGI, the report adds.
Efforts were analyzed in six categories:
risk assessment;
current damage;
security structures;
existential security;
governance and transparency, and;
information sharing.
Pope warns about artificial intelligence
No company received an "A", the highest grade, in any category, while Anthropic received the best overall grade, a "C+".
The report highlights that several companies that previously banned the military use of their technology have been "gradually reversing course", including Anthropic, which was criticized for maintaining "questionable military commitments".
The United States government used artificial intelligence technology. Anthropic in military operations in Venezuela and Iran over the past year, according to multiple press reports, although the company was recently the subject of a Pentagon ban over disagreements over AI safety.
While "there are constructive attempts," the efforts as a whole are "totally insufficient." Other risks include the possible misuse of a model to carry out a cyberattack or perform tasks potentially harmful to humans.

Source: G1