Japanese social media firms set in-house rules on AI use

Some Japanese media companies are embracing generative artificial intelligence to raise their productivity. But they are also setting workplace rules to prevent data leaks and privacy infringements.

Online media platform Note has been encouraging its staff to use ChatGPT for certain tasks. One example is writing headlines for its online magazine.

But it has classified information in its database into four levels of sensitivity. Workers are banned in principle from entering the most sensitive data into the program.

The head of Note's legal compliance office says balance is important when using ChatGPT. "We will utilize AI when we think it is useful as long as we can manage sensitive information properly."

Social-network provider Mixi has adopted a type of generative AI that does not use protected data to train the software.

Workers are allowed to input project plans, programming codes and other types of data held in-house as long as personal information is excluded.