Okta has introduced a new tool called Model Context Protocol or MCP scouting which aims to reduce the cost of running AI agents. According to Okta, this is achieved by listing each model call made by an AI agent and including schemas, names, descriptions and parameters for every tool exposed by a MCP server. This results in a significant reduction in AI agent token costs.
The term "tool tax" refers to the tokens consumed as a model considers tools, including these exposures to the MCP server. By analyzing each interaction with a tool, Okta can identify inefficiencies and optimize the process for running AI agents more cost-effectively.
This new tool is part of Okta's efforts to improve security in the use of artificial intelligence, where security measures must be carefully considered to ensure effective protection against cyber threats. With MCP scouting on board, businesses can expect further reductions in AI agent token costs, making it easier and more affordable to deploy AI-powered systems without breaking the bank.