Frontier models

Short description: Highly capable general-purpose AI models at the cutting edge of artificial intelligence technology

Frontier models are the most advanced artificial intelligence models that represent the cutting edge of AI capabilities at any given time. These highly capable foundation models push the boundaries of what is possible with AI technology, typically characterized by their massive scale, multimodal capabilities, and ability to perform a wide variety of complex tasks across different domains. The term encompasses both foundation models and general-purpose AI systems that exceed the capabilities of existing models and often require significant computational resources to develop and deploy, typically exceeding 10^25 floating-point operations (FLOPs) during training.

Definition and Characteristics

Core Definition

The concept of frontier models was formally introduced in a 2023 paper by OpenAI titled "Frontier AI Regulation: Managing Emerging Risks to Public Safety," where they are defined as "highly capable foundation models that could possess dangerous capabilities sufficient to pose severe risks to public safety." This definition highlights three key challenges: dangerous capabilities can emerge unexpectedly during development or post-deployment; it is difficult to prevent misuse of deployed models; and model capabilities can proliferate rapidly, especially through open-source releases.

The United Kingdom government, a key proponent of the term, defines frontier AI as "highly capable general-purpose AI models that can perform a wide variety of tasks and match or exceed the capabilities present in today's most advanced models." The UK's analysis emphasizes that frontier models are increasingly multimodal and may be augmented into autonomous AI agents with tool use (for example web browsing, code execution) that expand their real-world impact.

The Frontier Model Forum, an industry body established by leading AI companies, defines frontier models as "large-scale machine-learning models that exceed the capabilities currently present in the most advanced existing models, and can perform a wide variety of tasks."

Key Characteristics

Frontier models typically exhibit several defining characteristics:

Characteristic	Description	Implication
Massive Scale and Compute	Trained using vast computational resources, often exceeding 10^25 FLOPs, with parameter counts in the hundreds of billions or trillions, and training costs ranging from tens to hundreds of millions of dollars	High development costs create barriers to entry, concentrating power in well-funded labs. Scale is a primary driver of advanced capabilities
Multimodal Capabilities	Ability to process and generate multiple types of data including text, images, audio, and video	Enables broad applicability across domains but complicates safety evaluation and control
Emergent Properties	Display abilities that were not explicitly programmed, such as complex reasoning, chain-of-thought reasoning, code generation, or creative writing	Makes pre-deployment risk assessment challenging. The full capabilities may not be understood until widely deployed
General-Purpose Functionality	Designed to be adaptable across a wide range of domains and tasks with minimal fine-tuning	Enormous utility but difficult to predict all possible uses, including malicious applications
Extended Context Windows	Modern frontier models can process extensive amounts of information, with some models supporting context windows of over 2 million tokens	Enables complex, long-form tasks but increases potential for sophisticated manipulation
Autonomy Potential	Advanced models show increasing ability to act autonomously to achieve goals, using tools, accessing external information, and self-correcting	Raises long-term safety concerns about control and alignment with human values

History and Development

Evolution of the Term

The history of frontier models is intertwined with the evolution of foundation models, which emerged in the late 2010s. The term "foundation model" was coined in 2021 by researchers at the Stanford Institute for Human-Centered Artificial Intelligence (HAI) to describe large-scale models trained on broad data using self-supervision.

The first models trained on over 10^23 FLOPs were AlphaGo Master and AlphaGo Zero, developed by DeepMind and published in 2017. The progression to current frontier models included:

Timeline of Major Developments

Date	Milestone	Significance
2017	AlphaGo Master/Zero	First models exceeding 10^23 FLOPs
2018	BERT (Google)	Demonstrated power of transformer-based language models
2019	GPT-2 (OpenAI)	Raised concerns about misuse, leading to staged release
2020	GPT-3 (OpenAI)	175 billion parameters, sparked widespread interest in large language models
March 2023	GPT-4 (OpenAI)	First model exceeding 10^25 FLOPs, demonstrated significant capability leap
May 2023	Joint AI Safety Statement	CEOs of major AI labs declare extinction risk from AI a global priority
July 2023	Frontier Model Forum launched	Industry coordination body established by Anthropic, Google, Microsoft, and OpenAI
October 2023	U.S. Executive Order 14110	Established compute-based reporting requirements for frontier models
November 2023	Bletchley Declaration	28 countries agree on frontier AI risks and cooperation
2024	EU AI Act finalized	Establishes regulations for "general-purpose AI models with systemic risk"
2024-2025	Proliferation of frontier models	Over 30 models exceed 10^25 FLOP threshold

Current Examples of Frontier Models

As of 2025, several models are considered to be at the frontier of AI capabilities:

Model Name	Developer	Release Date	Key Capabilities	Training Compute (Estimated)	Parameter Count
GPT-4	OpenAI	March 2023	Multimodal processing, advanced reasoning, code generation	~2×10^25 FLOPs	Undisclosed (est. 1.76 trillion)
GPT-4o	OpenAI	May 2024	Optimized multimodal, real-time voice and vision capabilities	~10^26 FLOPs	Undisclosed
Claude 3 Opus	Anthropic	March 2024	Extended context (200K tokens), strong reasoning, constitutional AI approach	>10^25 FLOPs	Undisclosed
Claude 3.5 Sonnet	Anthropic	June 2024	Complex reasoning, analysis, creative tasks, computer use capabilities	>10^25 FLOPs	Undisclosed
Gemini Ultra	Google DeepMind	December 2023	Natively multimodal, strong benchmark performance	>10^25 FLOPs	Undisclosed
Gemini 1.5 Pro	Google DeepMind	February 2024	Massive context window (up to 2 million tokens in preview)	>10^25 FLOPs	Undisclosed
Llama 3.1 405B	Meta	July 2024	Open-source, multilingual, long-context reasoning	>10^25 FLOPs	405 billion
Grok-3	xAI	2025	Real-time knowledge integration, truth-seeking emphasis	Undisclosed	Undisclosed
DeepSeek-V2	DeepSeek	2024	Cost-efficient, strong in coding and math	>10^25 FLOPs	236 billion

Capabilities and Applications

Performance Benchmarks

Leading frontier models achieve near or above human-level performance on various standardized benchmarks:

Benchmark	Description	Frontier Model Performance	Human Performance
MMLU	Multitask Language Understanding	86-90%	~89%
GSM8K	Grade School Math	90-95%	~90%
HumanEval	Code Generation	85-90% pass rate	Variable (expert: ~95%)
GPQA	Graduate-level Science Questions	50-60%	PhD experts: ~65%
BigBench-Hard	Challenging reasoning tasks	80-85%	~85%

Application Domains

Domain	Capabilities	Example Applications	Key Models
Natural Language Processing	Text understanding, generation, translation, summarization	Chatbots, content creation, document analysis, code generation	All frontier models
Computer Vision	Image/video analysis, generation, manipulation, understanding	Medical imaging, autonomous vehicles, creative tools, quality inspection	GPT-4V, Gemini, Claude 3
Scientific Research	Data analysis, hypothesis generation, pattern recognition, simulation	Drug discovery, climate modeling, genomics research, materials science	GPT-4, Claude 3 Opus
Code Generation	Writing, debugging, explaining, and optimizing complex code	Software development, automation, technical documentation, DevOps	GPT-4, Claude 3.5 Sonnet
Multimodal Processing	Integration of text, image, audio, video inputs/outputs	Virtual assistants, content moderation, accessibility tools	GPT-4o, Gemini 1.5
Autonomous Agents	Tool use, web browsing, API calls, multi-step reasoning	Research assistants, customer service, workflow automation	Claude 3.5 (computer use), GPT-4 (plugins)

Regulatory Frameworks

European Union AI Act

The European Union AI Act establishes specific regulations for what it terms "general-purpose AI models with systemic risk," which closely aligns with the concept of frontier models.

Aspect	Requirement	Details
Compute Threshold	10^25 FLOPs	Models exceeding this threshold are presumed to have systemic risk
Model Evaluations	Mandatory testing	Standardized protocols and adversarial testing required
Risk Assessment	Systemic risk evaluation	Must assess and mitigate potential societal-scale risks
Incident Reporting	Report to AI Office	Serious incidents must be reported to EU AI Office
Cybersecurity	Adequate protections	Ensure model and weight protection against misuse
Documentation	Technical documentation	Comprehensive documentation for downstream providers
Compliance Deadline	August 2025	Full implementation required for covered models

United States Framework

The U.S. approach centers on Executive Order 14110 on Safe, Secure, and Trustworthy AI:

Component	Threshold	Requirements
Dual-use Foundation Models	>10^26 FLOPs	Report to government; share safety test results
Biological Sequence Models	>10^23 FLOPs (if primarily biological data)	Enhanced scrutiny for biosecurity risks
Red Team Testing	All covered models	Required before deployment
NIST AI RMF	Voluntary framework	Risk management guidance for AI lifecycle
AI Safety Institute	Established 2023	Develops standards and evaluation frameworks

United Kingdom Approach

The UK focuses on capability-based assessment rather than rigid compute thresholds:

• AI Safety Institute (renamed AI Security Institute in 2025): Evaluates models across four categories:

1. Safeguards effectiveness
2. Autonomy capabilities
3. Human influence potential
4. Societal resilience

• Pre-deployment Testing: Voluntary agreements with major labs for model access
• International Leadership: Hosts AI Safety Summits and coordinates global governance efforts

California Legislation

California has taken a leading role in U.S. state-level frontier AI governance:

• SB 1047 (Failed, 2024): Would have required safety testing for models >10^26 FLOPs and >$100M training cost; vetoed by Governor Newsom
• SB 53 (Passed, 2025): Transparency in Frontier AI Act establishing reporting mechanisms and whistleblower protections

International Frameworks

The OECD updated its AI Principles in 2024 to address frontier models, emphasizing:

• Risk-based approach to regulation
• International cooperation on standards
• Focus on trustworthy AI development

Safety and Governance Initiatives

Frontier Model Forum

Established in July 2023 by Anthropic, Google, Microsoft, and OpenAI, the Frontier Model Forum serves as the primary industry coordination body.

Aspect	Details
Current Members (2025)	Amazon, Anthropic, Google, Meta, Microsoft, OpenAI
Key Objectives	AI safety research, best practices, policy collaboration, societal applications
AI Safety Fund	$10 million for independent safety research
Executive Director	Chris Meserole (appointed October 2023)
Focus Areas	CBRN risks, cyber capabilities, societal impacts, evaluation standards

AI Safety Institute International Network

Launched at the AI Seoul Summit in May 2024, creating global collaboration:

Member Country	Institute Status	Key Focus
United States (Chair)	NIST USAISI	Standards and evaluation frameworks
United Kingdom	AI Security Institute	Pre-deployment testing and evaluation
European Union	AI Office	Regulatory compliance and enforcement
Japan	AI Safety Institute	International standards alignment
Singapore	AI Verify Foundation	Testing and certification
Others	Various stages	Canada, France, Kenya, Australia, South Korea

Bletchley Declaration

Signed by 28 countries in November 2023, establishing international consensus on:

• Recognition of frontier AI's transformative potential and risks
• Need for safety testing and evaluation standards
• Commitment to international cooperation
• Human-centric, trustworthy AI development

Risks and Challenges

Technical and Capability Risks

Risk Category	Description	Examples	Mitigation Approaches
Emergent Capabilities	Unexpected abilities appearing at scale	In-context learning, tool use, deception	Comprehensive evaluation, capability discovery research
Hallucination	Generation of false/misleading information	Fabricated citations, incorrect facts	Improved training, retrieval augmentation, uncertainty quantification
Jailbreaking	Bypassing safety constraints	Harmful content generation, misuse instructions	Adversarial training, constitutional AI, robust safety layers
Loss of Control	Models acting beyond intended parameters	Reward hacking, mesa-optimization	Alignment research, interpretability, killswitches

Misuse Risks

• Cybersecurity Threats: Automated vulnerability discovery, sophisticated phishing, malware generation
• Information Warfare: Large-scale disinformation campaigns, deepfakes, propaganda
• CBRN Risks: Lowering barriers to chemical, biological, radiological, or nuclear weapon development
• Privacy Violations: Personal data extraction, surveillance capabilities, profiling

Societal and Structural Risks

Risk	Impact	Affected Groups
Labor Displacement	Automation of cognitive work	Knowledge workers, creative professionals
Economic Concentration	Market dominance by few companies	Smaller firms, developing nations
Bias Amplification	Perpetuation of historical prejudices	Marginalized communities
Democratic Erosion	Manipulation of public discourse	Citizens, democratic institutions
Environmental Impact	Massive energy consumption	Global climate, local communities

Resource Requirements and Development Costs

Computational Requirements

Aspect	Current Scale (2025)	Projected (2027)
Training Compute	10^25 - 10^26 FLOPs	10^27 - 10^28 FLOPs
Training Cost	$100M - $500M	$1B - $5B
Training Duration	3-6 months	6-12 months
GPU Requirements	10,000 - 50,000 GPUs	100,000+ GPUs
Energy Consumption	50-200 GWh	500+ GWh

Infrastructure Needs

• Hardware: Specialized GPUs (H100, H200) or TPUs
• Data Centers: Hyperscale facilities with advanced cooling
• Networking: High-bandwidth interconnects (InfiniBand)
• Storage: Petabyte-scale distributed systems
• Software Stack: Custom training frameworks and optimization tools

Future Directions

Technical Developments

• Scaling Projections: Models may reach 10^28 FLOPs by 2027-2028
• Efficiency Improvements: Focus on reducing inference costs while maintaining capabilities
• Enhanced Multimodality: Integration of additional modalities (3D, robotics, biological data)
• Agent Capabilities: More sophisticated tool use and autonomous operation
• Reasoning Advances: Improved logical reasoning and planning capabilities

Research Priorities

According to the International Scientific Report on Advanced AI Safety:

• Mechanistic interpretability: Understanding model internals
• Scalable oversight: Human supervision of advanced systems
• Adversarial robustness: Resistance to attacks and manipulation
• Alignment research: Ensuring models pursue intended goals
• Capability evaluation: Better benchmarks and testing methodologies

Governance Evolution

• Development of international standards through ISO/IEC
• Implementation of EU AI Act provisions (August 2025)
• Potential U.S. federal legislation on AI safety
• Expansion of AI Safety Institute network
• Industry self-governance through Frontier Model Forum

See Also

• Foundation models
• General-purpose AI
• AI safety
• Large language models
• Transformer architecture
• AI alignment
• Compute governance
• Artificial general intelligence
• AI risk management
• Existential risk from artificial general intelligence
• AI Safety Summit