Terrecon, a Ukrainian nonprofit, identified 981 AI-generated songs promoting pro-Russian war propaganda on platforms like Spotify and YouTube, with most released after 2024 due to accessible AI music tools; the songs praise Russian forces, use ethnic slurs, and depict the war favorably, raising concerns about content moderation and the weaponization of AI in information warfare.
AI-generated summary
Terrecon, a Ukrainian threat intelligence nonprofit, conducted research into AI-generated music related to the Russia-Ukraine war, identifying nearly 1,000 tracks with pro-Russian propaganda content across major streaming platforms.
Among the songs whose release dates could be established, most appeared after 2024. (Reuters photo)
Artificial intelligence is being used to produce a growing number of songs about the Russia-Ukraine war, with some of the music presenting Russia's invasion in a favourable light and criticising Ukraine. Hundreds of such tracks have appeared on major music streaming platforms, including Spotify, Apple Music, YouTube and Deezer, according to research by Terrecon, a Ukrainian nonprofit organisation focused on threat intelligence. The songs cover a wide range of musical styles, from ballads to heavy metal. Terrecon said some praise Russian soldiers, describe the war from a Russian perspective and contain derogatory references to Ukrainians, including ethnic slurs. One of the more popular tracks identified by the organisation is тАЬPower of ArtilleryтАЭ, a Russian-language rock song centred on artillery systems including Krasnopol and Carnation.
Nearly 1,000 songs identified
Terrecon said it catalogued 981 songs created with AI that contained what it classified as pro-Russian war propaganda. Among the songs whose release dates could be established, most appeared after 2024. That coincided with the wider availability of AI music generators such as Sudo and Udio. The technology has lowered the time and cost involved in producing music, according to Daria Verbytska, a founder of Terrecon. тАЬThey have found new ways to create cheaper drones and use them on a massive scale,тАЭ Daria Verbytska, a founder of Terrecon, said told NYT in a telephone interview, referring to the Russians. тАЬItтАЩs like that with music тАФ creating massive propaganda with the help of AIтАЭ Some of the songs use themes associated with Soviet-era war music but present them through contemporary genres. One track, тАЬSoldier's DaughterтАЭ, describes the death of a soldier and the impact on his child while also portraying the soldier's sacrifice in a positive light. It is not clear how much involvement, if any, the Russian government has had in producing or distributing the songs. However, the Kremlin has supported cultural projects aimed at promoting patriotic themes. Last week, a director of Moscow's Internet Development Institute, an organisation that works closely with the Kremlin to promote patriotic values, said in a presentation that the institute had supported more than 350 artistic projects related to the war, according to Russian media reports.
AI tools can produce war-related music
Terrecon also tested several AI music generators to see how they responded to prompts requesting songs promoting Russian views of the war. Researchers used Suno, Udio and Google Flow Music in the US, along with Singapore-based Mureka. They entered prompts in both Russian and English, including requests involving violence and derogatory descriptions. The platforms rejected some prompts, but researchers said they were nevertheless able to generate songs resembling some of those in Terrecon's database. They then uploaded the tracks to YouTube Music and SoundCloud. тАЬThe moderation systems mainly responded to specific banned words rather than the overall meaning of the prompts,тАЭ TerreconтАЩs report found. The researchers said the issue is partly linked to the differing rules adopted by AI music services and streaming platforms. They also said automated systems may have difficulty identifying expressions, slang and offensive language in different languages.
Streaming platforms tighten AI rules
Major music platforms generally have policies covering hate speech and violent content, although fewer have specific rules dealing with propaganda. Deezer and YouTube are among the platforms that require certain AI-generated content to be labelled. Spotify said in September that it had introduced a policy to label AI-generated songs and prevent them from being recommended to users. The company also added labels to songs shared by The New York Times as examples. Terrecon said the songs it identified had been played millions of times across major streaming services, although the extent to which listeners understood that some of the music was AI-generated or politically motivated is unclear. Deezer spokesman Jesper Wendel said the platform reviews potentially rule-breaking content on a case-by-case basis. SoundCloud spokeswoman Sade Ayodele said the company reviews reported content and added that тАЬthere is no place on SoundCloud for content that incites or glorifies violence.тАЭ The emergence of these songs illustrates how AI music tools can be used to produce and distribute large volumes of war-related material. It also presents streaming and AI companies with the challenge of deciding how existing rules on hate speech, violence, political content and AI-generated material should apply to such music.
AI outlook тАФ possibilities, not facts
Streaming platforms will introduce stricter labeling requirements for AI-generated political content
Likely ┬╖ Within months
AI music services will face increased pressure to improve detection of propaganda and hate speech in multiple languages
Likely ┬╖ Within months
A six-month live facial recognition trial by British Transport Police at London railway stations cost ┬г320,786 and scanned over 500,000 faces, producing only one watchlist alert тАФ which was a false positive with no resulting arrest. Despite the low accuracy, BTP extended the programme, sparking renewed debate over cost, privacy, and proportionality of biometric surveillance in public transport.
Google has launched the Fitbit Air in India, a screenless fitness tracker priced at тВ╣13,999 with 24/7 heart rate, sleep, and SpO2 monitoring. It integrates with Google Health app and comes with a three-month trial of Google Health Premium. Sales begin October 2 via Google Store and retail partners. The device competes with Noise REP Band, Amazfit Helio Strap, and Whoop in India's growing screenless wearable market.
President Donald Trump signed an executive order directing federal agencies to replace the terms 'Artificial Intelligence' and 'AI' with 'Super Intelligence' and 'SI' in official communications, citing the technology's evolution beyond its original definition. The order follows a White House luncheon with tech leaders including Elon Musk, Mark Zuckerberg, Sundar Pichai, and Jensen Huang, and mandates a 60-day deadline for proposing legislation to establish a new statutory definition of 'super intelligence'.
The Supreme Court of India has urged the Centre to legally enforce an 18-year age threshold for social media accounts, agreeing that contracts with minors are void and calling for statutory mechanisms to prevent underage access, citing risks like online grooming and data misuse.

In 1933, Percy Shaw saw light coming from a cat's eye on the road, leading to the invention of the reflective stud 'cat's eyes', which is still used around the world today; Now scientists are creating high-tech cameras inspired by the eyes of cats and other animals, which could be useful in robots, drones and autonomous vehicles.
SpaceX's Starship rocket is set to attempt its first orbital flight from Texas, aiming to deploy 26 Starlink satellites, orbit Earth six times, and land in the Pacific Ocean, supporting future Moon and Mars missions despite inherent risks.