Open Access
Review
Table 6.
Information security countermeasures at AI-empowered decision layer
| Threats | Countermeasures | Remarks | Representative solutions |
|---|---|---|---|
| Secure aggregation algorithm | Federated learning | [138, 229–231] | |
| Combined with blockchain | Federated learning | [232, 233] | |
| Insecure and malicious AI | Security hardware | Federated learning | [234] |
| model behaviors | Reinforcement models | Federated learning | [235] |
| Security architecture | Large language model | [236–238] | |
| Security training | Large language model | [239–243] | |
| Security inference | Large language model | [244–246] | |
| Secure Multi-Party Computing | Federated learning | [247] | |
| Differential Privacy | Federated learning | [248, 249] | |
| User data privacy | Homomorphic encryption | Federated learning | [25, 250] |
| leakage | Partial parameter sharing | Federated learning | [251, 252] |
| Corpus cleaning | Large language model | [253, 254] | |
| Privacy enhancing technology | Large language model | [255–257] | |
Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.
Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.
Initial download of the metrics may take a while.