Seven in every 100 adults surveyed in Australia, the United Kingdom and the United States say someone has used AI to make or spread sexual images of them without their consent. The survey of over 7,200 people, led by Nicola Henry of RMIT University in Australia and published in the International Journal of Human–Computer Interaction, counted three kinds of abuse since age 18. These were having a nude or sexual image of oneself created or altered with AI tools, having such an image shared, and being threatened with its release. Men reported it at over twice the rate of women, 10.2% against 3.8%, and rates were also higher among people under 35, people from racial and ethnic minorities and people with disabilities.

Women reported more fear for their safety and more damage to their mental and physical health. Almost 87% of all victims described at least one harm to their health, relationships, work or finances. Henry warned that the real number may be higher, because some people never find out that a fake image of them exists. The research was funded by Google, and a Google researcher, Rebecca Umbach, co-wrote it. Henry points to Australia’s proposed digital duty of care, which would require online services to prevent foreseeable harm instead of reacting after it happens. She also asks for better education, accessible support for victims and for technology companies to build safeguards into AI image tools from the start.

Schools are being encouraged to adopt AI tools partly on the strength of research reviews that report large gains in learning, and a new audit argues those reviews do not hold up. Patrick O’Neill, an associate professor at Ivy Tech Community College in Indiana, examined 14 meta-analyses, studies that combine the results of many smaller studies into one overall estimate. In a preprint discussed by Education Week on November 2, he concludes that none of them gives a valid basis for its claims. None clearly defined the kind of AI use it was measuring, and none properly checked for publication bias, the tendency for studies with good results to be published while disappointing ones are left out.

Some of the problems were basic. A study of an elementary school physical education class was grouped with one on simulations for medical students, a few included studies did not use AI at all, and results as different as test knowledge, motivation and mental health were added together. Some of the underlying studies contained impossible statistics or references that appear to be invented, including ChatGPT research dated before the tool was public. The reviews appeared in respected journals and have been cited thousands of times. O’Neill asks journals to require authors to share their data and to tighten peer review. A separate Stanford review of more than 800 studies found only 20 with a strong cause-and-effect design, and none of those involved schoolchildren.

Spotify is aiming to launch a paid tool in early 2027 that would let subscribers turn existing songs into AI-made covers and remixes, such as a rock track reworked as a dance version, the Financial Times reported on October 30. The add-on could cost an extra $4 to $5 a month on top of Premium, though the price has not been decided and Spotify has not confirmed the plans. According to the report, progress has been slower than expected because major current stars have been harder to sign up than some established older artists. Universal, which first announced a remix deal with Spotify in May, has joined, as have the independent-label agency Merlin and the distributor Believe, while Sony and Warner are still negotiating.

Artists signed to those labels can still decide for themselves whether their songs are used, and those who take part are expected to receive a share of the extra fees when fans alter their music. Olivia Rodrigo told The Hollywood Reporter in early October that she would never use AI in her music. Bernstein analyst Ian Moore estimates that 7 to 8 percent of Spotify’s paying U.S. customers could use the service by the end of 2028. Before any full launch, co-CEO Gustav Söderström has said, a public test version will open to some users.