Explore the 2025 CCF‑Ant Research Fund: 50 Cutting‑Edge Projects in AI, Security & Computing

Fund Overview

CCF‑Ant Research Fund (first batch 2025) is officially launched, with the application deadline of August 25 2025 24:00 (Beijing time). Since its inception in 2020, the fund has supported over 150 research projects with more than 50 million CNY in total, covering fields such as artificial intelligence, privacy computing, architecture, databases, and blockchain.

Fund Themes and Topics

Direction 1: Data Security and Privacy Protection

Four frontier topics focus on encrypted computation, post‑quantum security, container security, etc.

Real‑time detection and analysis of transaction security on blockchain

AI model architectures friendly to encrypted computation

Post‑quantum secure protocol design and analysis

Next‑generation container security architecture and intelligent protection

Direction 2: Hardware‑Software Co‑Design

Fifteen topics address homomorphic encryption performance, OS innovation, formal verification, and more.

Ciphertext inference network based on SNN

ANT‑ACE large‑input data partitioning and layout optimization

Code generation, optimization, and scheduling for multi‑accelerator systems

High‑performance modular multiplication design

High‑performance NTT architecture design

KeySwitch hardware accelerator

Heterogeneous homomorphic computing system based on compute‑storage devices

Rack‑level heterogeneous computing system for homomorphic encryption applications

Silicon‑pre‑verification and system‑level evaluation tools for dedicated homomorphic encryption chips

Secure containers and isolation based on RustOS

Undefined‑behavior detection for RustOS using Miri

Automatic translation or generation of RustOS drivers with LLM

Modular verification framework for concurrent Rust operating systems

Automated generation and evaluation of Rust code contracts for secure OS

Reliability verification of embodied intelligent agents in HW‑SW co‑design

Direction 3: Supercomputing and Intelligent Computing

Five topics target high‑performance computing, large‑model training/inference, and inference optimization.

Distributed memory semantic communication

Large‑scale shared infrastructure architecture, performance and stability research

Resilient and hot‑migration capabilities for large‑model training/inference collective communication

Dynamic continuous profiling‑driven large‑model inference optimization

Power‑aware high‑performance near‑data computing scheduling framework

Direction 4: Artificial Intelligence

Twenty‑six topics cover data & AI, AI security, domain models, and intelligent agents.

Intelligent web content extraction research

Long‑context and RAG inference acceleration based on high‑performance KVCache storage

Semantic‑driven cross‑scenario recommendation system based on large language models

GUI Agent applications in customer service

Risk ranking contrastive learning for auto‑insurance pricing

Dynamic proxy selection and multi‑agent collaboration in intelligent agent service platforms

Support for low‑resource languages in international finance large models

Reinforcement‑learning‑based liquidity layout decision system

Hidden risk research in open credit services

Multimodal reasoning for external data fusion

Warehouse‑level program and logic consistency verification

RAG system evaluation and monitoring optimization

Generative recommendation unified modeling paradigm

VLA control instruction edge generation system

Medical imaging large model construction with low‑annotation dependence

Specialized clinical reasoning large model based on few‑shot learning and multimodal fusion

User debt planning large model

Real‑time fund demand perception and dynamic intelligent service decision system based on multimodal large models

Multi‑modal experience diagnosis large model based on user behavior understanding

Personalized large model research based on global behavior data

Multi‑modal reinforcement learning for UI intelligent agents

AI face‑swap and high‑definition screen liveness attack‑defense research

Knowledge mining and intelligent agent reasoning enhancement in closed‑domain self‑evolution

Domain graph construction for graph‑retrieval‑enhanced generation technology

Hierarchical memory system graph‑structure enhancement for agents

Multi‑agent collaborative optimization based on reinforcement learning

Eligibility

Full‑time faculty or researcher at a global university/research institute (post‑docs excluded)

CCF members are given priority when qualifications are equal

Applicants must be able to conduct independent research and lead student teams

Application Deadline and Submission

The deadline is August 25 2025 24:00 (Beijing time). Applications should be submitted through the Ant Group Technology Research Institute website by uploading the proposal. A QR code for the submission portal is provided below.