For 20 years, ISD has delivered field-leading threat detection, analysis and real-world strategies to combat terrorism, extremism and authoritarianism - in all their ideological forms.

Recommender Systems and Algorithms

ISD’s research explores how algorithms and generative AI can amplify terrorist content as well as weaponised hate and influence operations across online platforms. This offers valuable insights into potential risks to democratic processes and extremist indoctrination. Our advisory work on social media algorithms explores solutions to these challenges including changes to policies and platform designs, as well as greater transparency through mechanisms including algorithmic audits.

November 12, 2026

TikTok and the mainstreaming of the True Crime Community

Violent TCC content is reaching wider audiences ever since users migrated from Tumblr to TikTok, coinciding with a sharp rise in linked attacks.

Hybrid Extremism and Nihilistic Violence, Kids Online Safety, Recommender Systems and Algorithms, Tech Accountability and Safety, Threat Detection and Analysis

July 22, 2026

Radicalisation in Closed Loops: Risks and Intervention Opportunities with AI Chatbots and Companions

Recommender Systems and Algorithms, Tech Accountability and Safety, Tech Legislation and Regulation, Threat Analysis and Prevention

May 5, 2026

Amplifying Antisemitism: How Recommender Algorithms Serve Harmful Content to Children

Antisemitism, Recommender Systems and Algorithms, Tech Accountability and Safety

May 2, 2026

Recommendation systems and the rapid exposure to antisemitism online

Antisemitism, Recommender Systems and Algorithms, Tech Accountability and Safety

January 20, 2026

Men, algorithms and how YouTube shapes media consumption

Recommender Systems and Algorithms, Tech Accountability and Safety

October 27, 2025

Popular chatbots are citing Russian state propaganda

Recommender Systems and Algorithms

July 11, 2025

AI blurs reality amid Iran-Israel conflict

Emerging Technologies, Information Warfare and Online Manipulation, Recommender Systems and Algorithms

June 12, 2025

ISD highlights information operations trends and AI use amid LA protests

Foreign Information Manipulation and Interference, Recommender Systems and Algorithms

May 20, 2025

Recruited by algorithms and AI: Terrorists in the West are getting younger

Recommender Systems and Algorithms, Terrorism and Extremism

Further Reading

January 20, 2025

ISD Written Evidence to the Science, Innovation and Technology Committee Inquiry, on Social Media, Misinformation and Harmful Algorithms

This document examines the harms of social media algorithms, their role in the 2024 UK riots, and regulatory approaches to mitigate misinformation and

Recommender Systems and Algorithms, Tech Legislation and Regulation

November 9, 2024

Disconnected from reality: American voters grapple with AI and flawed OSINT strategies

An ISD analysis ahead of the US elections found users misidentified AI content in over 50% of cases in a research sample, often labeling real content

Democratic Integrity, Recommender Systems and Algorithms

October 30, 2024

ISD’s Elise Thomas for Lawfare: Gen AI will increase misinformation about disinformation

Elise Thomas, OSINT Senior Analyst at ISD, wrote an op-ed for Lawfare on how generative AI amplifies misinformation.

Emerging Technologies, Recommender Systems and Algorithms

October 8, 2024

The Irish Channel: Investigating an Irish Misinformation Hub, Political Connections and AI Hallucinations

The report examines the Irish Channel website and social media, a key source of misinformation in Ireland.

Anti-Migrant, Democratic Integrity, Far-Right Extremism, Recommender Systems and Algorithms, Targeted Threats, Hate and Abuse, Terrorism and Extremism

September 19, 2024

Platforms fail to label and remove AI-generated and manipulated election content

With the 2024 US election rapidly approaching, ISD has found evidence that platforms are failing to remove or label even the debunked content.

Democratic Integrity, Emerging Technologies, Recommender Systems and Algorithms, Threat Analysis and Prevention

June 19, 2024

Pulling Back the Curtain: An Exploration of YouTube’s Recommendation Algorithm

This is the executive summary of ISD’s Pulling Back the Curtain series which explores different aspects of YouTube’s recommendation algorithm.

Information Warfare and Online Manipulation, Kids Online Safety, Recommender Systems and Algorithms

June 18, 2024

New research points to failures by YouTube in protecting children from harmful content

A four-part ISD investigation reveals misogynistic content, and videos related to self-harm and suicide, are being recommended to young users.

Recommender Systems and Algorithms, Tech Accountability and Safety

June 18, 2024

Investigation 1: Teenage Gamer Personas

This is the first investigation in ISD’s Pulling Back the Curtain series which explores different aspects of YouTube’s recommendation algorithm. The series consists of four separate investigations examining the recommendations sent to eight U.S.-based YouTube accounts registered with varying ages, genders, and interests over one month. This report examines the recommendations sent to two teenager accounts with an interest in gaming.

Gaming and Alternative Platforms, Recommender Systems and Algorithms, Tech Accountability and Safety

June 18, 2024

Investigation 2: “Male Lifestyle Guru” Personas

This is the second investigation in ISD’s Pulling Back the Curtain series which explores different aspects of YouTube’s recommendation algorithm. The series consists of four separate investigations examining the recommendations sent to eight U.S.-based YouTube accounts registered with varying ages, genders, and interests over one month. This report examines the recommendations sent to two accounts with an interest in "male lifestyle gurus”.

Misogyny, Recommender Systems and Algorithms, Tech Accountability and Safety

June 18, 2024

Investigation 3: Mommy Vlogger Personas

This is the third investigation in ISD’s Pulling Back the Curtain series which explores different aspects of YouTube’s recommendation algorithm. The series consists of four separate investigations examining the recommendations sent to eight U.S.-based YouTube accounts registered with varying ages, genders, and interests over one month. This report examines the recommendations sent to two accounts with an interest in Mommy vloggers.

Recommender Systems and Algorithms, Tech Accountability and Safety

Previous 12345 Next