This work deals with cooperation of Legacy Information Systems. The growth and variety of distributed information sources imply a need to exchange and/or to share information extracted from various and heterogeneous databases. Cooperation of legacy information systems requires advanced architectures able to solve conflicts coming from data heterogeneity. The project ACSIS (Agents for Cooperation of Secure Information Systems) is a solution to resolve semantic conflicts coming from databases heterogeneity taking care of scalability issues. Therefore, this solution uses Distributed Artificial Intelligence techniques like defined Interaction Protocols and evolutive domain ontology.