AI-assisted violence is pushing women out of public life, finds a new UN Women report co-authored by researchers at City St George’s, University of London. More women in public life are self-censoring online and at work to avoid abuse, unwanted sexual advances in DMs, or deepfakes. The study surveyed 641 women in public life – journalists, activists, human rights defenders – in late 2025. The full report with breakdowns per category can be downloaded here: https://ow.ly/NlYy50YRS2i.

Topline stats:

- 27% received unsolicited sexual advances via direct message

- 12% had personal images (sometimes intimate) shared without their consent

- 6% were subjected to deepfakes or manipulated images and videos

- 24% experienced anxiety or depression as a result

- 41% self-censored on social media to avoid harassment. 

New research from the University of Oxford suggests that losing a parent in adulthood can have lasting consequences for earnings, through deteriorating mental health and reduced access to family support, with earnings effects particularly marked for women and for mothers with young children. 

A new non-invasive brain stimulation technique known as accelerated continuous theta burst stimulation (a-cTBS) improves social communication at one month follow up and has a favourable safety profile in children with autism, finds a trial from China published by The BMJ today. 

Study examines how survivors of stealthing - nonconsensual removal of a condom by an intimate partner - process and communicate the experience to confidants and health care workers. A traumatic experience for survivors.

In a significant breakthrough for adolescent mental health in India, The George Institute for Global Health India, has successfully demonstrated a scalable, community‑based care model through Project ARTEMIS—the country’s first large‑scale study focused on adolescents living in urban slums—showing measurable reductions in depression, self‑harm and suicide risk across 60 slum communities in Delhi and Vijayawada.

Researchers from the University of Kansas' school of journalism and mass communications analyzed journalism class syllabi from 15 universities to learn more about how they address the usage of AI. They found varying approaches, from banning its use and considering it academic dishonesty, to allowing approved use or treating it as a subject of professional inquiry. The varying approaches can cause confusion and shortchange students, while a more consistent policy could be helpful, the researchers argue.