In the last week of September 2026, Turkish courts jailed two more journalists under Article 217/A of the Penal Code, the 2022 "disinformation" offence. According to the Committee to Protect Journalists, Ali Tarakçı, a YouTube broadcaster, was taken from his home around midnight on September 25 and, the next day, was arrested pending trial by Ankara's 8th Criminal Court. The trigger was an X post on September 23 saying "a very important minister will resign on Friday." His X account has also been blocked in Turkey. Fatih Ergin, news coordinator of the daily Nefes, was arrested on September 28 over a post about the death of a doctor, Turgay Tamer. CPJ says more than a dozen journalists have been detained or arrested under the law since January 2026.
The case for the law
The strongest argument for an offence like this is real. Viral falsehoods during disasters, outbreaks or security incidents can cause panic, and a state has a legitimate interest in deterring deliberate fabrication. Turkish lawmakers also argued that the provision was narrowly drawn. The Constitutional Court's majority, as reported by MLSA, read it to require that the speaker act knowingly with intent to create fear or panic, and that the information be capable of disturbing public peace. If the law worked that way, it would be a fraud-style statute aimed at deliberate deceivers.
What the statute actually does
The text is broader than that reading. Article 217/A was added by Law No. 7418 in October 2022 and carries one to three years in prison, according to the Turkish Penal Code on mevzuat.gov.tr. The Council of Europe's Venice Commission, in its urgent opinion on the draft, recommended that Turkey not enact it. Its concerns were that the wording was too ambiguous to meet the legality test and that the chilling effect made it doubtful the offence was necessary or proportionate in a democratic society. Parliament passed it anyway.
The Constitutional Court then upheld the provision, but not unanimously. MLSA reports six members dissented, including then-President Zühtü Arslan. He warned that the "almost the only clear and predictable aspect" of the clause was the prison sentence. Another dissenting member argued that a three-year maximum is a disproportionate intervention in freedom of expression.
The Tarakçı case shows the problem. A prediction that a minister will resign is a claim about the future. It cannot be a verifiable falsehood when it is posted, and it is hard to see how it could create public panic. If the prediction was wrong, that is a reputational problem for the speaker, not a public-order emergency. Even on the Constitutional Court's own reading, the prosecution has to show knowing falsity and intent to cause fear. Pretrial detention is imposed before either has been tested.
Error is not deception
CPJ's account of Ergin's case describes a post that, in the authorities' view, misidentified a doctor's role in a death linked to a reopened investigation. That is an alleged factual error in a fast-moving story. Journalists make such errors, and the normal remedies are correction, right of reply, and civil liability for defamation. One dissenting judge made this point directly: it is hard for journalists to verify developing news instantly. A criminal offence that cannot tell the difference between a mistake and a lie pushes reporters toward silence on exactly the contested stories where scrutiny matters most.
The breadth of the problem matters more than any single case. In April 2026, CPJ and 22 other organisations called the law "structurally incompatible with international press freedom standards" and named earlier cases, including İsmail Arı of BirGün, detained on March 21, 2026. CPJ now says authorities promised in 2022 that the law would not be used against journalists. A dozen-plus detentions in nine months is not a series of isolated misapplications. It is what a vaguely worded offence with a short, discretionary path to detention produces.
A proportionate alternative
A pro-innovation, pro-speech approach does not mean ignoring harmful falsehoods. It means choosing tools that scale to the harm:
- Target the harm, not the falsehood. Fraud, defamation, incitement and election-interference offences already exist in Turkish law and require identifiable victims or concrete harms.
- Put a court review before detention. Speech cases should not start with custody on the strength of a single post, and the courts should say why detention is necessary.
- Restrict blocking to specific content. Blocking an entire X account, as happened to Tarakçı, removes lawful speech along with whatever was disputed, and it is a heavy remedy for a post about a minister's future.
- Use platform transparency over criminal liability. Labels, community notes and correction mechanisms address misinformation without turning users into defendants.
The Venice Commission's 2022 warning was specific: ambiguous wording invites arbitrary use. Four years of enforcement against journalists have borne that out. Turkey can still take the Commission's advice. Repealing 217/A, or at minimum narrowing it to demonstrable, knowing falsehoods that cause concrete harm, would retain the deterrent against real fabrication while ending detentions over predictions and honest mistakes. Meanwhile the platforms operating in Turkey face a recurring question about how far to comply with account-blocking orders that follow from these cases. Whatever they decide, the underlying fix has to come from the statute.