AI for Charities - How to Manage the Threat of AI Slop

This AI for charities guide on how to manage the risk of AI slop is one of our best practice guides that form part of our free Charity AI Ready programme to support charities in safely and effectively adopting AI.

What is AI Slop?

AI slop is a relatively new phrase, somewhat akin to an online version of e mail spam.  A wave of mass-produced AI generated content that is beginning to flood blogs, social media, and search results. It's low-quality content, which might not be harmful but is of little value.  It can be text, images, video, and even entire websites.

What Does AI Slop Do?

These AI content farms generate slop usually to boost SEO and cash in on advertising revenue.  The aim is not to mislead but to convince or to sound confident.  A definition I like is careless speech - AI  generated output that contains “subtle inaccuracies, oversimplifications or biased responses that are passed off as truth in a confident tone,”  (Source: Royal Society: Do large language models have a legal duty to tell the truth? Aug 24).  Or, if you prefer,  the moral philosopher Harry Frankfurt’s definition of ‘bullshit’, which is when a speaker is

.....unconcerned with whether or not they’re saying the truth – it’s irrelevant as their goal is persuasion. Liars at least acknowledge that it matters what is true. ....bullshit is a greater enemy of the truth than lies are.

Misinformation and disinformation on the other hand is false or inaccurate information spread intentionally or unintentionally.  You can use our AI Fact Check Resource to deal with misinformation.

Is AI Slop a Risk to Charities?

Slop is not just another online irritation.  Trust is fundamental to charities, not only to ensure people trusts us enough top use our services but also to donate to fund those services.  AI slop can make it more difficult for your charity to stand out to reach those people but it can also damage trust in your charity too.

How to Spot Slop

The first step in dealing with slop is to identify it.

• Repetitive Content: The same information repeated, including across platforms.  I find that the kind of slop I've had and blocked all sounds very similar.
• Lack of Context: Content that is out of place or lacks relevant details.
• Poor Quality: Grammatical errors, awkward phrasing, or irrelevant information.

What AI Slop Looks and Sounds Like

Here's an example blocked from my own site, trying to drive traffic from our small charity Christmas Market web page.  It's from a an unusual source I don't recognise that claims to be a German web design site.

Hi! I just finished reading your article, and I have to say it was really informative! You did a great job of covering points that people often overlook, and I loved how clearly you explained everything. This really struck a chord with me, and I think many others will find it helpful as well. I’ve been exploring a related topic on my site, where I discuss [mention related subject]. I’d love to hear what you think about it! Thanks again for sharing such great content. Keep up the awesome work!

Practical Steps to Protect Your Charity From AI Slop

• Automated Moderation: Implement moderation tools to filter out low-quality content- see below.  Personally, I've simply blocked all comments on our website.
• Block Slop Specific Words: Most social media platforms allow you to create a list of blocked words. Comments containing these words will be automatically hidden or flagged for review.  For example:
  •   On Facebook, go to Page Settings > General > Page Moderation to add blocked words.
  •   On Instagram, go to Settings > Privacy > Comments to add custom keyword filters.
• Human Moderation: Ensure that you social media and feeds are routinely monitored and any slop blocked.
• Community Reporting:  Regularly engage with your audience by liking, replying, and encouraging meaningful interactions. This helps to promote quality discussions and reduces the likelihood of AI slop.  Encourage your followers to report any inappropriate content. Use the block feature to prevent repeat offenders from spamming your feeds.
• Regular Audits: Conduct regular audits of your content to ensure quality.  For example, decide how often you will perform content audits, allocate responsibility for doing so and develop a checklist which might include the relevance of content, quality of writing (grammar, spelling, clarity), identifying and blocking slop specific words, accuracy of information, presence of outdated content and consistency in tone and style.

AI slop is different to misinformation but you might combine the 2 in your audit, as there is a lot of overlap.  Here's our Fact Checker Best Practice Guide.

Tools for Managing AI Slop

We are likely to see platforms becoming increasingly able to block slop. For example, Google has already deployed Perspective API.  It’s a free tool that uses machine learning to help identify and filter toxic comments and content, making it useful for moderating social media and chat feeds.  We are also likely to see more from Social Media Management systems.  For example, Hootsuite offers AI-powered moderation tools that can help small charities manage their social media content and filter out unwanted or low-quality posts.

The kinds of developments we're likely to see are:

• Advanced AI Filters: Improved AI algorithms to detect and remove slop.
• User Education: Platforms educating users on identifying and reporting slop.
• Collaboration with Fact-Checkers: Partnerships with fact-checking organizations to verify content.

Hopefully, in due course, slop will end up similar to span.  Something that needs to be dealt with but which is, in large measure, dealt with automatically.

AI for Charities - Trust is Critical 

We think that a combination of growing AI risks, such as slop, misinformation and scams pose a significant and increasing threat to the trust with our beneficiaries, donors and the public that is so critical to our work.

High Quality Comms.  AI slop is low quality, repetitive and often irrelevant.  Consequently, a very good way to protect your charity is by ensuring your own content is everything it isn't, so people can more easily recognise it as the rubbish it is.

• Quality Over Quantity: Focus on creating high-quality, valuable content.
• Engage with Your Audience: Build a strong relationship with your audience to foster trust.
• Transparency: Be clear about your sources and intentions.

Although not specific to AIU slop, here are 2 other ways in which you could help to maintain trust in your charity.

Existing Communications Guidelines.  Clearly state your community guidelines and posting rules in the About section of your social media profiles. This sets expectations for your audience and deters low-quality comments.  For example, something along the lines of

Please be respectful and relevant in your comments. Inappropriate or off-topic comments will be removed.  Serious or repeated breaches will result in a user being blocked. 

What Are Your Charity AI Policies?  Ensure that your charity has clear policies on AI use that will make sure your use of AI is transparent, trustworthy and of value to your audience, including posts and comments in your feeds and anything that you quote or repost.  You can download 60 charity policies from within Charity Excellence, which have been updated to reflect AI where appropriate and which include a simple AI policy, and our AI Ready Programme web page connects you to free AI guides and training to help you.

AI Slop - UK Regulation and Law 

We have the Online Safety Act 2023 and UK GDPR, we have the Advertising Standards Authority (ASA) and the Government is working on regulations to address AI-generated content but I cannot find anything that specifically relates to slop.