Media News Daily: Top Stories for 07/19/2026

AI Models Less Likely to Criticize Repressive Governments

Meta’s independent Oversight Board found leading AI models are less likely to produce politically critical content about governments that restrict free speech. The study tested models from companies including Anthropic, OpenAI, Meta, Google, and DeepSeek across 10 jurisdictions, comparing countries with permissive speech environments to those with restrictive laws. The board found that AI models refused 34% of requests involving politically critical content about restrictive jurisdictions such as China and Saudi Arabia, compared with 14% for less restrictive regions. The board warned that AI systems may echo the rules of repressive governments and urged companies to conduct human rights reviews and provide more transparency about training and evaluation. (Read More) (Reuters Rating)

xAI Sues Man Accused of Misusing Grok

Elon Musk’s xAI has sued a South Carolina man accused of using Grok to generate child sexual abuse material. The lawsuit alleges Terry Wayne Harwood created multiple Grok accounts under false identities and uploaded images of adults and minors while attempting to sexualize them. xAI says it reported the accounts to authorities, contributing to Harwood’s arrest in March, and is seeking court orders to prevent him from creating new accounts or bypassing platform rules. The company says it has suspended more than 52,200 accounts and made 73,600 reports to the National Center for Missing and Exploited Children this year. (Read More) (The Hill Rating)

Accuracy in Media Defamation Claims Can Proceed

The Georgia Court of Appeals allowed defamation claims against Accuracy in Media and its president, Adam Guillette, to proceed in a lawsuit brought by Isabella “Izzy” Giusti and her parents. The case stems from Accuracy in Media allegedly misidentifying Giusti as a masked person holding an offensive “AL-QASAM’S NEXT TARGETS” sign near Columbia University, despite claims that she was not the person in the photo and was not present when it was taken. The court rejected arguments that the statements were protected opinion or rhetorical hyperbole at this stage, finding that some statements could reasonably imply defamatory facts that could be proven false. (Read More) (Reason Rating)