International Report Warns of Growing Risks as Artificial Intelligence Capabilities Accelerate

International Report Warns of Growing Risks as Artificial Intelligence Capabilities Accelerate
———————————————
An international annual report on artificial intelligence (AI) safety has warned of escalating risks associated with the rapid expansion of AI technologies, particularly in areas such as digital fraud, deepfake content, cyberattacks, regulatory challenges, and potential impacts on the labor market.

The report, described as a comprehensive global survey of AI-related technological developments and risks, was led by Canadian computer scientist Yoshua Bengio and prepared with contributions from more than 100 independent experts. It is part of a series launched following the International AI Safety Summit in 2023, according to The Guardian.

While noting significant advances in AI performance in mathematics, science, programming, and image generation, the report highlighted persistent shortcomings, including the production of inaccurate information, known as “hallucinations,” and the inability of current systems to reliably carry out long, complex tasks independently.

Despite these limitations, cited studies indicate rapid improvement in software engineering capabilities, with the duration of tasks AI systems can perform reportedly doubling every seven months. If this trend continues, AI could handle multi-hour tasks by 2027 and multi-day tasks by 2030, raising concerns about job displacement, even though full automation of complex projects remains out of reach for now.

The report also flagged the spread of AI-generated deepfake pornography as a serious concern and warned of AI’s growing role in cyberattacks. On labor markets, it found mixed evidence of impact, noting uneven adoption across countries and sectors, and concluded that more profound employment effects may emerge as AI systems become more capable and autonomous.