From optional to essential: How AI is changing research at CSSR

By: Derrick Sekgala, Zea Leon, Jane Kelly, Rebecca Maughan-Brown, and Elona Toska.

The CSSR research team, led by Derrick Sekgala, launched a study to understand how researchers are using AI tools, the opportunities they are finding, and the challenges they face. While AI use in higher education is growing rapidly, little was known about how CSSR researchers themselves were engaging with these tools, or what guidance and support they needed.

| “AI is no longer optional; it has become part of the standard research toolkit.” - Derrick Sekgala.

One striking finding: 96% of respondents reported using AI tools, even though most rated their own understanding as only basic or moderate. This shows a critical reality: researchers are using AI extensively, but often without formal training or deep technical knowledge. AI is now integrated into standard research workflows, not just an optional extra.

Why ChatGPT dominates among AI tools

ChatGPT is the most widely used. Its early visibility, ease of use, and free access have made it familiar and approachable, leading many researchers to adopt it as their default tool.

The risks researchers are concerned about

Despite high adoption, researchers remain cautious. Key concerns include: 

• Accuracy
• Ethical challenges and academic integrity
• Lack of clear institutional guidance
• Uneven understanding of appropriate use

Researchers are engaging thoughtfully with AI, but often without frameworks to navigate risks confidently.

Why training has lagged

Many researchers have received no formal AI training, largely because adoption outpaced universities’ ability to develop programmes. Most learning has been self-directed rather than institutionally supported.

What researchers want 

Researchers emphasise the need for practical, hands-on guidance that supports their daily work. Key priorities include:

• Scenario-based workshops
• Clear ethical guidelines and best practices
• Technical support for selecting and implementing tools
• Peer learning networks Discipline-specific training

The bottom line in one sentence

CSSR researchers are using AI widely, creatively, and critically, but without formal training or institutional support, there is an urgent need for structured guidance to ensure responsible, ethical, and effective use.

Read the full working paper here.