I’ve just caught up on last week’s workshop run by the charity, BEAT, about eating disorders and social media.
One word that stood out in the panel discussions was ‘nuance’. This topic, which also expanded out artificial intelligence (AI), is full of nuance – it’s complex, ever evolving and deeply human.
Below I’ll outline some key takeaways from the workshop and their implications for people in recovery and those supporting them.
What’s helpful for one, is unhelpful for another
Social media and AI can be both helpful and unhelpful depending on the user. For one person a lifeline and a pillar of recovery, for another deeply triggering and reinforcing harmful mindset and behaviours.
For people with eating disorders, who are often isolated, sometimes geographically, sometimes because of the shame and secrecy that so often accompanies these conditions, online communities can offer something of real value: lived experience, peer support and a sense of being understood. Recovery stories and honest conversations shared in comment sections can be the thing that helps someone feel less alone.
However, the same platforms that help facilitate those connections also serve up content shaped by algorithms that don’t distinguish between what’s helpful and what’s harmful. For someone who isn’t in a good place, a feed that mirrors their anxieties back at them can reinforce exactly what they’re already thinking, pull their attention away from their meal plans, fuel comparison, and make it harder to communicate clearly with the medical professionals trying to support them.
This isn’t just a young person’s issue
The discussions highlighted something which I think is so valuable to keep in mind with eating disorders. They don’t discriminate by age, gender, background, or where you live. And this follows through in terms of the use of social media.
The assumption that social media’s influence on eating disorders is primarily a teenage problem is not borne out by evidence coming out of recent studies. Adults of all ages can be affected, particularly when their health is already vulnerable and they’re spending more time online. When someone is in a vulnerable mental state, they may not realise the negative impact that what they’re consuming is having.
A survey by First Steps ED found 87% of respondents encountered harmful content daily and 62% didn’t know how to report it. Therefore, it was good to hear from an OFCOM representative about measures being put in place make online spaces safer.
The impact of rapidly evolving AI
It was very valuable that the workshop also discussed the impact AI is having. I know from my counselling practice that many people are using AI for support and information about health issues. AI is increasingly part of people’s emotional lives, and for people with eating disorders, that can raise some challenging questions.
AI can provide users with what is being termed ‘unconditional empathy’ which can feel validating. But, of course, validation isn’t always the same as psychological and ethical support. A doctor spoke about cases where AI has reinforced unhelpful beliefs, deepened emotional dependencies, or left people less likely to turn to medical professionals who might actually challenge what they’re experiencing.
As AI rapidly evolves, the ethical implications of its use for people with eating issues are still becoming apparent. One recent study found that all the chatbots researched served up harmful content1 . Speakers at the workshop highlighted that it’s too early for much research about the impacts to be available. More research and an ongoing dialogue between users and those around them are much needed so safeguards and training can be developed.
What can help
Despite the nuances and rapid pace of evolution in social media and AI, there are some aspects are practical and helpful.
Social media literacy is important, for both individuals and those who support them. This can involve helping people understand how algorithms work, that images may be manipulated, and how to curate feeds that are more varied and useful. It can also be helpful for individuals to identify what content they find triggering. Supporting them reconnect to different interests can help to broaden the content of their feed.
As ever, lived experience and honest portrayal of how recovery is presented is valuable. Recovery is typically not linear. So, educating people about the hopeful, whilst ‘messy’ reality of living with an eating disorder can be so helpful. Non-judgmental spaces, online and offline, where people can speak freely, are worth protecting and building.
Importantly, this isn’t something any one group can solve alone. Healthcare professionals, people with lived experience, carers, charities, tech companies, and regulators all need to be in conversation — ideally co-developing resources with the people most affected. First Steps ED also spoke about something I found reassuring: face-to-face community support still matters. Digital connection is powerful, but it doesn’t replace everything.
Final summary
What I took away from this workshop is that whilst there’s a lot more to be done to develop safe environments for people with eating disorders who interact with social media and AI, there’s a lot of positive value these platforms can provide. It’s important to acknowledge there’s complexity in eating disorders, social media and AI. But also that everyone with an eating disorder has their own unique journey to recovery and technology can be helpful.
If you’re supporting someone with an eating disorder, it’s valuable to engage in conversations about social media and AI with genuine curiosity. This can help you navigate and understand it together. The UK’s leading eating disorders charities, BEAT and First Steps ED, also provide valuable resources and support and I’ve included some examples below.
Helpful advice links from BEAT:
Eating disorders and social media
Posting and viewing eating disorder recovery content on social media
Supporting someone with an eating disorder
First Steps ED’s guide to staying safe online – Scroll Safe: A Student’s Guide to Online Wellbeing
Ofcom’s online safety super-complaints page
A useful video about this subject – Social media, food and me
1 Yim, S. H., Yoo, D. W., Polymerou, A., Liu, Y., & Saha, K. (2025). Generative AI for Eating Disorders: Linguistic Comparison With Online Support and Qualitative Analysis of Harms. International Journal of Eating Disorders. https://doi.org/10.1002/eat.24604
