New research supported by over 30 nations, as well as representatives from the EU and the UN, shows the impact AI could have if governments and wider society fail to deepen their collaboration on AI safety.
The ‘International Scientific Report on the Safety of Advanced AI’ is an interim report developed by a diverse global team of AI experts, which aims to give policymakers across the globe a single source of information to inform their approaches to AI safety. It also provides businesses and organizations with a guide to the latest thinking on emerging AI risks and potential impacts.
The report recognises that advanced AI can be used to boost wellbeing, prosperity, and new scientific breakthroughs, but it notes that like all powerful technologies, current and future developments could result in harm. The report highlights a lack of universal agreement among AI experts on a range of topics, including both the state of current AI capabilities and how these could evolve over time. It also explores the differing opinions on the likelihood of extreme risks which could impact society such as large-scale unemployment, AI-enabled terrorism, and a loss of control over the technology.
The report summaries the challenges of managing risks associated with advanced AI as follows:
- Developers still understand little about how their general-purpose AI models operate.This is because general-purpose AI models are not programmed in the traditional sense. Instead, they are trained: AI developers set up a training process that involves a lot of data, and the outcome of that training process is the general-purpose AI model. These models can consist of trillions of components, called parameters, and most of their inner workings are inscrutable, including to the model developers.
- General-purpose AI is mainly assessed through testing the model or system on various inputs. These spot checks are helpful for assessing strengths and weaknesses, including vulnerabilities and potentially harmful capabilities, but do not provide quantitative safety guarantees. The tests often miss hazards and overestimate or underestimate capabilities because general-purpose AI systems may behave differently in different circumstances, with different users, or with additional adjustments to their components.
- Independent actors can, in principle, audit general-purpose AI models or systems developed by a company. However, companies often do not provide independent auditors with the necessary level of direct access to models or the information about data and methods used that are needed for rigorous assessment.
- It is difficult to assess the downstream societal impact of a general-purpose AI system because research into risk assessment has not been sufficient to produce rigorous and comprehensive assessment methodologies. In addition, general-purpose AI has a wide range of use cases, which are often not predefined and only lightly restricted, complicating risk assessment further.
