Artificial Intelligence and the Next Generation of Domestic Abuse
SAFECHAIN™ · Digital Safeguarding
Artificial Intelligence and the Next Generation of Domestic Abuse
By Samantha Avril-Andreassen, LLB (Hons), LLM, LPC, FRSA
Founder, SAFECHAIN™
Artificial intelligence is transforming the way we work, communicate and make decisions.
It can summarise information, translate languages, detect fraud, automate routine tasks and improve access to services. Used responsibly, AI has enormous potential to improve people's lives.
However, history shows that every significant technological advancement carries a parallel question:
How might it be misused?
For safeguarding professionals, that question has become increasingly urgent.
Artificial intelligence is not creating domestic abuse.
People do.
What AI is doing is providing new tools that can make coercive control more convincing, more scalable and, in some cases, more difficult to detect.
The challenge is no longer simply understanding domestic abuse.
It is understanding how abuse evolves alongside technology.
Technology Changes. Coercive Control Adapts.
The fundamental dynamics of coercive control remain unchanged.
It is still about power.
It is still about domination.
It is still about restricting another person's freedom.
Artificial intelligence simply introduces new ways of achieving those same objectives.
As organisations embrace AI, safeguarding frameworks must evolve at the same pace.
Emerging Risks
Artificial intelligence can be exploited in ways that extend existing patterns of abuse.
Potential examples include:
generating convincing fake messages or emails;
creating synthetic voice recordings that imitate another person;
producing manipulated images or videos designed to intimidate or discredit;
automating harassment across multiple digital platforms;
impersonating trusted individuals through AI-generated conversations;
using AI to gather, organise and analyse personal information about someone more efficiently.
Many of these technologies have entirely legitimate uses.
The safeguarding challenge lies in recognising when they are deliberately used to frighten, isolate, manipulate or control another person.
When Reality Can No Longer Be Trusted
One of the most significant consequences of generative AI is the erosion of certainty.
If a convincing recording can be fabricated...
If a realistic photograph can be generated...
If messages can be produced that appear authentic...
How do professionals distinguish genuine evidence from manipulated content?
For survivors, this creates an additional burden.
They may be required not only to disclose abuse, but also to defend the authenticity of the evidence they present.
At the same time, perpetrators may deny genuine evidence by claiming it was "created by AI."
Safeguarding systems must therefore prepare for both possibilities:
fabricated evidence; and
false claims that authentic evidence has been fabricated.
Psychological Manipulation at Scale
Artificial intelligence also changes the speed at which coercive behaviour can occur.
Tasks that previously required significant effort may become automated.
Repeated messages.
Persistent monitoring.
Impersonation.
Image manipulation.
Coordinated online harassment.
Although the technology changes, the psychological objective remains familiar:
To create confusion.
To undermine confidence.
To isolate.
To control.
The Institutional Challenge
Most safeguarding frameworks were developed before widespread public access to generative AI.
Consequently, many organisations are still asking questions designed for earlier forms of digital abuse.
Current safeguarding assessments may not ask:
Was AI used to impersonate someone?
Could digital evidence have been manipulated?
Has synthetic media been used to intimidate the victim?
Has AI enabled harassment across multiple platforms?
Does the organisation have expertise to assess emerging technologies?
Without these questions, important risks may remain invisible.
Institutional Fragmentation™
Artificial intelligence introduces challenges that cross organisational boundaries.
Police may investigate digital offences.
Technology companies may hold platform data.
Banks may identify unusual account activity.
Schools may notice behavioural changes.
Healthcare professionals may observe psychological harm.
Domestic abuse services may understand the coercive pattern.
Each organisation holds one part of the picture.
Few hold the whole.
This is Institutional Fragmentation™ in action.
The technology evolves across sectors faster than safeguarding systems evolve between them.
AI Must Strengthen Safeguarding—Not Undermine It
Artificial intelligence is increasingly being embedded into public services.
It may assist with:
risk assessment;
case prioritisation;
document review;
customer support;
fraud detection;
administrative decision-making.
These developments offer opportunities.
But safeguarding cannot become dependent upon automated decisions alone.
Human judgement, professional curiosity and trauma-informed practice remain essential.
AI should support safeguarding professionals.
It should never replace professional accountability.
Questions Every Organisation Should Be Asking
As AI becomes more widely adopted, organisations should ask:
Does our safeguarding policy address AI-enabled abuse?
Are frontline professionals trained to recognise synthetic media and digital impersonation?
Can staff distinguish between cyber incidents and safeguarding disclosures?
Are AI systems independently governed and regularly reviewed for safety?
Do we have clear procedures for assessing AI-generated evidence?
Are survivors informed about emerging digital risks?
If the answer is "not yet", there is an opportunity to prepare before these issues become commonplace.
The SAFECHAIN™ Perspective
Artificial intelligence is neither inherently safe nor inherently dangerous.
Like every technology before it, its impact depends upon how people choose to use it—and how institutions choose to govern it.
SAFECHAIN™ views AI-enabled domestic abuse as an emerging safeguarding challenge requiring collaboration between technology companies, policymakers, regulators, public services and specialist domestic abuse organisations.
The future of safeguarding will not depend solely upon developing more intelligent technology.
It will depend upon building more intelligent systems of governance.
Because artificial intelligence should never become another instrument of coercive control.
It should become part of the solution.
This article reflects SAFECHAIN™'s analysis of emerging safeguarding risks associated with artificial intelligence and technology-facilitated domestic abuse. It is intended to support professional discussion and does not constitute legal, cybersecurity or regulatory advice.
SAFECHAIN™
Building safeguarding systems that recognise patterns—not just incidents.
Copyright
© 2026 Samantha Avril-Andreassen. All Rights Reserved.
Published by SAFECHAIN™
This publication is protected by international copyright law. No part of this article may be reproduced, distributed, transmitted, stored in a retrieval system, translated, adapted or published in any form or by any means without the prior written permission of the copyright holder, except for brief quotations used for academic, educational, journalistic or review purposes with appropriate attribution.
SAFECHAIN™, Institutional Fragmentation™, Participation Integrity™, Disclosure Integrity™, Jurisdictional Integrity™, The Cumulative Harm Model™, Quick Exit™, Stealth Mode™, Journal Lock™, and all associated SAFECHAIN™ frameworks, methodologies, models, standards and publications are proprietary intellectual property of Samantha Avril-Andreassen.
This article is published to promote professional discussion, safeguarding awareness and systems reform. It does not constitute legal, cybersecurity, regulatory or professional advice. Readers should seek specialist advice appropriate to their circumstances.