Heads of state in Europe, North America and Japan will be invited to sign an international treaty banning hate speech online at a meeting of the Council of Europe next January.

The document, an additional protocol to the Convention on Cybercrime that was ratified last November by members of the Council of Europe, seeks to criminalise "acts of a racist and xenophobic nature conducted through computer systems".

The aims of the protocol are to harmonise criminal law and to improve international co-operation in fighting racism and xenophobia.

It defines racist and xenophobic material as written material, images or other representations of ideas or theories advocating, promoting or inciting hatred, discrimination or violence against individuals or groups, based on race, colour, descent, national or ethnic origin, or religion if used as a pretext for any of these factors.

It calls on signatories to adopt legislative and other measures to make it a criminal offence to distribute or make available such material to the public through a computer system. Distribution would cover the sending of email containing such material, while the creation of web pages or even links to web pages containing such material would be considered "making available".

The preamble to the protocol expresses the beliefs that "acts of a racist and xenophobic nature constitute a violation of human rights" and that "national and international law need to provide adequate legal responses to propaganda of a racist and xenophobic nature through computer systems", while "recognising that freedom of expression constitutes one of the essential foundations of a democratic society".

The protocol is intended only to cover intentional acts which means, for example, that ISPs would not be held liable for hosting pages or transmitting email messages containing racist or xenophobic material, as long as they were just serving as a conduit for the material.