frontier AI
In plain wordsThe most capable general-purpose AI systems of the moment — those at or beyond the leading edge — which governments single out for extra safety attention.
AI at the leading edge of capability. In the working definition the UK government adopted for the 2023 AI Safety SummitAI Safety Summit, prop. n.: A series of international meetings of governments, AI companies and researchers, begun in England in 2023, to agree on how to handle the risks and benefits of the most advanced AI., it means highly capable general-purpose models, usually foundation modelsfoundation model, n.: A very large AI model, built once at great cost from vast amounts of text, images or code gathered mostly from the internet, that many products then reuse as their starting point, from chatbots to coding tools., that can handle a wide range of tasks at or beyond the level of the most advanced models of the day; such systems may have dangerous capabilities that have not yet been observed. The category is relative and moves with the state of the art, which is both its purpose — aiming new rules at new risks without burdening ordinary AI — and its main difficulty for lawmakers. In practice it is operationalized through proxies: training-compute thresholdscompute threshold, n.: A legal cutoff based on how much computation was used to train an AI model. Models trained with more than that amount face extra rules, on the theory that bigger training runs produce more capable models. (such as the EU AI Act's presumption at 10²⁵ FLOPFLOP, n.: A single arithmetic step, such as multiplying or adding two numbers. Counting them measures how much computing went into an AI model, and laws use such counts to decide which models face extra rules. for systemic-risk modelsgeneral-purpose AI model with systemic risk, n.: The EU's label for the very largest and most capable AI models, whose failure or misuse could cause harm across society, so their makers must test them for dangers and report serious incidents.), capability thresholdscapability threshold, n.: A level of dangerous ability — such as meaningfully helping someone make a biological weapon — that, once an AI model reaches it, obliges its developer to add stronger safeguards or hold the model back. identified by dangerous capability evaluationsdangerous capability evaluation, n.: A test of whether an AI system can do things that could cause severe harm in the wrong hands or on its own — such as helping make a bioweapon, hacking computers or copying itself., or lists of named developers.
The institute tests frontier AI before release under voluntary agreements with the labs.
Origin
Popularized in July 2023 by the paper 'Frontier AI Regulation: Managing Emerging Risks to Public Safety' (Markus Anderljung, Joslyn Barnhart and co-authors), by the industry Frontier Model Forum founded the same month, and by the UK government, which renamed its Foundation Model Taskforce the Frontier AI Taskforce in September 2023 and made frontier AI the focus of the Bletchley summit.
Usage note
frontier modelfrontier model, n.: One of the most capable AI models in existence at a given time, built by a handful of leading labs at great cost, and the main target of AI safety rules. names the systems and frontier labfrontier lab, n.: One of the handful of companies building the most advanced AI models of the moment, spending billions of dollars on computing power to train them. their developers. Critics argue that the framing, promoted largely by industry and safety researchers, shifts attention from present harms to speculative ones; the EU AI ActEU AI Act, prop. n.: The European Union's law on artificial intelligence, the first broad AI law anywhere: it bans some uses, sets strict checks for risky ones and puts duties on the makers of the biggest AI models. avoids the term, using 'general-purpose AI model with systemic risk' instead.
Related
- frontier modelfrontier model, n.: One of the most capable AI models in existence at a given time, built by a handful of leading labs at great cost, and the main target of AI safety rules.
- frontier labfrontier lab, n.: One of the handful of companies building the most advanced AI models of the moment, spending billions of dollars on computing power to train them.
- general-purpose AI model with systemic riskgeneral-purpose AI model with systemic risk, n.: The EU's label for the very largest and most capable AI models, whose failure or misuse could cause harm across society, so their makers must test them for dangers and report serious incidents.
- compute thresholdcompute threshold, n.: A legal cutoff based on how much computation was used to train an AI model. Models trained with more than that amount face extra rules, on the theory that bigger training runs produce more capable models.
- AI Safety SummitAI Safety Summit, prop. n.: A series of international meetings of governments, AI companies and researchers, begun in England in 2023, to agree on how to handle the risks and benefits of the most advanced AI.
- Frontier AI Safety CommitmentsFrontier AI Safety Commitments, prop. n.: Promises made in 2024 by major AI companies to publish safety plans for their most advanced models, including the danger points at which they would stop if the risks could not be controlled.
- dangerous capability evaluationdangerous capability evaluation, n.: A test of whether an AI system can do things that could cause severe harm in the wrong hands or on its own — such as helping make a bioweapon, hacking computers or copying itself.
Sources
- Anderljung, M., Barnhart, J. et al. (2023). Frontier AI Regulation: Managing Emerging Risks to Public Safety. arXiv:2307.03718.
Mentioned in
AI regulationAI regulation, n.: Laws and binding government rules about artificial intelligence: what AI may be used for, what checks its makers must run, and what penalties apply when they break the rules., AI Safety InstituteAI Safety Institute, n.: A government agency that tests the most advanced AI models for dangerous abilities and studies how to make them safe. Britain and the United States set up the first ones in 2023., AI Safety SummitAI Safety Summit, prop. n.: A series of international meetings of governments, AI companies and researchers, begun in England in 2023, to agree on how to handle the risks and benefits of the most advanced AI., frontier modelfrontier model, n.: One of the most capable AI models in existence at a given time, built by a handful of leading labs at great cost, and the main target of AI safety rules., Bletchley DeclarationBletchley Declaration, prop. n.: A short statement signed by 28 countries and the European Union at the first global AI safety summit, held in England in 2023, agreeing that the most advanced AI could cause serious harm., Frontier AI Safety CommitmentsFrontier AI Safety Commitments, prop. n.: Promises made in 2024 by major AI companies to publish safety plans for their most advanced models, including the danger points at which they would stop if the risks could not be controlled., frontier labfrontier lab, n.: One of the handful of companies building the most advanced AI models of the moment, spending billions of dollars on computing power to train them., G7 Hiroshima AI ProcessG7 Hiroshima AI Process, prop. n.: An effort begun by the G7 group of wealthy democracies in Japan in 2023 that sets voluntary rules for companies building the most advanced AI, which now report publicly on how they follow them., general-purpose AI model with systemic riskgeneral-purpose AI model with systemic risk, n.: The EU's label for the very largest and most capable AI models, whose failure or misuse could cause harm across society, so their makers must test them for dangers and report serious incidents., International AI Safety ReportInternational AI Safety Report, prop. n.: A large scientific review, led by the computer scientist Yoshua Bengio and backed by more than 30 countries, summing up what is known about what advanced AI can do and how it could go wrong., Centre for the Governance of AICentre for the Governance of AI, prop. n.: A research organization, based in Britain, that studies how governments and companies should handle the dangers of advanced AI and the large economic and political changes it could cause., Collingridge dilemmaCollingridge dilemma, n.: The trap that a new technology is easiest to control when little is known about its harms, but by the time the harms are clear it is too widespread to change easily., Frontier Model ForumFrontier Model Forum, prop. n.: An industry group set up in 2023 by leading AI companies to share safety research and best practices for the most advanced AI models.