This topic is the next phase of the work.
Artificial intelligence is increasingly shaping how online harm is created, amplified, and distributed. While AI systems provide significant benefits in communication, education, and content moderation, they also introduce new forms of risk, particularly for women and children.
This section outlines key categories of AI-driven online harm to support awareness, education, and early recognition.
AI systems can be used to generate or manipulate digital content, including:
Deepfake images and videos used for impersonation or harassment
AI-generated explicit or abusive content targeting individuals
Synthetic media used for defamation or reputational harm
Automated generation of misleading or offensive text content
These forms of content can spread quickly and may be difficult to distinguish from authentic material.
AI tools can enhance the effectiveness of online manipulation techniques, including:
Highly personalised phishing messages
AI-assisted grooming or impersonation in chat environment
Automated scam messages that adapt to user responses
Large-scale targeting using profiling and behavioural data
These techniques increase the credibility and scale of online deception.
AI systems can be used to scale abusive behaviour, such as:
Bot-driven harassment campaigns on social media platforms
Coordinated trolling using automated accounts
Repeated targeting of individuals through generated messages
Amplification of harmful or abusive content through recommendation systems
This can result in sustained and difficult-to-mitigate online abuse.
AI technologies can support identity-related harms, including:
Creation of fake identities using synthetic profiles and images
Impersonation of real individuals using generated content
Modification or distortion of personal images or videos
Unauthorized reconstruction of a person’s likeness
Such misuse can affect trust, safety, and personal security online.
AI-driven recommendation and ranking systems may unintentionally:
Amplify harmful or extremist content
Promote sensational or abusive material for engagement
Reinforce exposure to repeated harmful narratives
Create feedback loops that increase vulnerability to manipulation
This represents a systemic risk rather than direct user-driven abuse.
AI-driven harm introduces new challenges for online safety, including:
Reduced ability to verify authenticity of digital content
Increased speed and scale of harmful content dissemination
Difficulty in attribution of harmful actions
Higher psychological and emotional impact on victims
To support users, CyberSafe recommends:
Developing awareness of AI-generated content indicators
Verifying information across trusted sources before reacting or sharing
Reporting suspicious or manipulated content on platforms
Preserving evidence when encountering harmful material
Seeking support from trusted individuals or safeguarding channels
This section is intended for awareness and educational use within CyberSafe. It does not replace legal, psychological, or institutional guidance, but aims to improve understanding of emerging AI-related risks in online environments.