Disinformation, Elections and the Public Sphere

In one line

Regulating lies means appointing somebody to identify them, and the person appointed is usually the person with the most to gain from the identification.

In the wording a student can write in an exam: disinformation is false information disseminated with intent to deceive, and misinformation is false information disseminated without that intent; the difficulty in regulating either is that a restriction on falsehood requires an authority to determine falsity, so the question is who decides, and in Kunal Kamra v. Union of India the Bombay High Court struck down the 2023 amendment to rule 3(1)(b)(v) of the Intermediary Rules, which made a fact check unit of the Central Government the identifier of what is fake or false or misleading about any business of the Central Government, holding it violative of Articles 14, 19(1)(a) and 19(1)(g), ultra vires the parent Act, vague and overbroad, incapable of being saved by reading down, failing the test of proportionality, and producing a chilling effect qua an intermediary.

Why this is the hardest speech problem

Every other restriction in [Speech Online and Its Limits] is defined by a quality of the speech that a court can assess: whether it incites, whether it defames, whether it is obscene. Falsity is different in three ways.

Truth is not among the article 19(2) grounds. The eight grounds are sovereignty and integrity, security of the State, friendly relations with foreign States, public order, decency or morality, contempt of court, defamation, and incitement to an offence. Falsity as such is not one of them, so a restriction on speech merely because it is untrue has to be brought within another ground, usually public order.

Determining falsity requires a determiner. A court can decide whether particular words defame a particular person, but a rule addressed to the whole of public discourse cannot wait for a trial in every case, so it needs a body that decides quickly. That body then holds the power the rule creates.

And the speech that matters most is contested. The statements that do political damage are usually not simple falsehoods but characterisations, predictions, selective truths and disputed claims, on which the determination of falsity is itself a political act.

Add the technology and the problem changes scale. [The Information Society and Human Rights] records that speech once reached as far as the speaker could be heard, and that the limit has gone. Recommendation systems select what each person sees, so a falsehood is not merely available but delivered; synthetic media makes fabrication cheap; and the same profiling that section 9(3) of the Digital Personal Data Protection Act 2023 forbids for children is what allows a message to be aimed at the people most likely to act on it.