Safeguarding Risks of Learner AI Use

Generative AI is no longer a niche technology; it is a mainstream tool that many of your learners and apprentices are already using. While the focus has often been on academic integrity, providers must now urgently consider the distinct safeguarding risks that these powerful platforms present. A proactive and educational approach is essential to keep learners safe in this rapidly evolving digital landscape.

Understanding the New Risk Landscape

Your existing online safety protocols may not fully cover the novel challenges posed by generative AI. Leaders and safeguarding teams need to understand this new risk profile to ensure protective measures remain effective. Key areas of concern include:

• Data Privacy and Security: Learners may unwittingly share personal, sensitive, or confidential information when using prompts in public AI models. This could include UCAS statements, personal reflections for assignments, or details about their circumstances, creating a permanent digital footprint beyond their control.
• Harmful or Inappropriate Content: AI tools can generate biased, inaccurate, or distressing content. Learners might be exposed to radical or extreme views presented as fact, or receive harmful advice on issues related to well-being or health.
• Manipulation and Misinformation: The ease of creating realistic but fake images, audio, and text (deepfakes) introduces new avenues for bullying, harassment, and coercion. This can range from creating embarrassing images of peers to more serious forms of identity-based abuse.
• Well-being and Dependency: There is a growing risk of emotional or psychological dependency on AI 'companions' or chatbots. This can hinder the development of social skills, resilience, and coping mechanisms, potentially isolating learners from real-world support networks.

Updating Your Policies and Procedures

To address these risks, your provider's existing frameworks need to be reviewed and updated. This is not about creating more bureaucracy, but about ensuring your policies are clear, relevant, and fit for purpose.

• Review Your Acceptable Use Policy (AUP): Explicitly name generative AI and outline acceptable and unacceptable use. Be clear about the use of personal data and the creation of content involving others.
• Distinguish Safeguarding from Malpractice: While related, using AI to cheat on an essay is an academic misconduct issue. Using AI to create a deepfake image to bully a peer is a safeguarding issue. Your policies must help staff distinguish between the two.
• Update Online Safety Policies: Integrate AI-specific scenarios into your online safety policy and training. Ensure it reflects the risks of data privacy, misinformation, and AI-generated content.
• Check Reporting Mechanisms: Ensure learners and staff know how to report a concern related to AI use. Is it clear that this falls under your safeguarding procedures? Is the process simple and accessible?

Educating Learners on Safe AI Engagement

Banning AI tools is unrealistic and counterproductive. Instead, the most effective safeguarding strategy is education. Your curriculum for participation and development should equip learners with the knowledge and skills to navigate AI safely and critically.

• Integrate AI Literacy: Build specific sessions on safe and ethical AI use into tutorials, induction programmes, and relevant curriculum areas.
• Promote Critical Consumption: Teach learners to question AI-generated content. Who created the tool? What are its biases? How can you verify the information?
• Use Practical 'Do and Don't' Scenarios: Move beyond abstract rules. Provide concrete examples: 'Do use AI to brainstorm ideas for your project; Don't paste your entire draft containing personal details into it.'
• Discuss Digital Citizenship: Broaden the conversation to include the ethics of AI, such as the impact of creating fake content, copyright implications, and how AI can perpetuate societal biases.

Equipping Staff for Confident Conversations

A vigilant safeguarding culture relies on all staff feeling confident to identify and respond to risks. Your team needs support to understand this new area.

• Provide Practical CPD: Offer training that goes beyond warnings. Show staff what these tools look like, how learners are using them, and what concerning output looks like.
• Include AI in Safeguarding Briefings: Make AI a regular agenda item in safeguarding updates, team meetings, and supervision sessions to normalise the conversation.
• Empower Tutors and Assessors: Equip frontline staff with the language and confidence to have supportive conversations with learners about their AI use, framing it around safety and digital skills rather than just detection and punishment.
• Prepare Designated Safeguarding Leads (DSLs): Ensure your safeguarding team understands how to triage, record, and respond to AI-related concerns, including when to escalate or involve external agencies.

Where this fits in QualityHero

Managing emerging risks requires robust systems. Within the QualityHero platform, the Safeguarding module provides a secure place to log, manage, and analyse AI-related concerns, helping you spot trends and evidence your response. You can use the Toolkit Areas module to audit your policies and curriculum against these new risks, creating and tracking actions in your QIP to ensure your safeguarding practices remain effective and responsive.