How I use AI tools to enhance EFL students’ speaking skills

Many university students studying English as a foreign language (EFL) face challenges with comprehension and communication. This could be because their secondary education was conducted in their mother tongue, coupled with limited opportunities to use English during their learning, both of which can lower students’ confidence and willingness to speak. I have turned to AI to enhance EFL students’ speaking abilities and boost confidence, and the results are promising.  

Self-directed AI speaking activity

A lack of confidence in their speaking fluency and accuracy is a significant hurdle for my students. To better prepare them, I have integrated GenAI tools, such as DeepSeek, with NaturalReader (a text-to-speech programme) to facilitate self-directed practice. Here’s how I use these tools to design accessible speaking tasks:

• Step 1. Before the start of each unit, I compile a glossary of key terms.
• Step 2. Students copy and paste the words in the glossary into NaturalReader.
• Step 3. I ask students to use GenAI to create passages that incorporate the glossary terms. For example, prompts such as the following work well: “I am an EFL student learning new vocabulary: [insert terms]. Create three short academic paragraphs (150 to 200 words) using these words in the context of [theme of the unit]”.
• Step 4. Students paste the AI-generated paragraphs into NaturalReader, listen to the audio and practise reading aloud. If the text is too complex, they can use GenAI to align the complexity to the Common European Framework of Reference for Languages (CEFR) A2/B1 level while retaining the key glossary terms.

By combining GenAI with AI voicing tools, students can practice mimicking speech at their own pace. This helps them articulate their ideas clearly and confidently during class discussions.

Personalised feedback with Microsoft Teams for Education

Many students miss out on personalised guidance because of large class sizes and limited class time. To offer timely and targeted feedback, I use Microsoft Teams for Education’s learning accelerator (Reading Progress) to prompt self-directed speaking practice and collect students’ performance data. Here’s how you can do the same:

• Step 1. Use GenAI to prepare materials for students to read aloud using prompts such as, “I am an EFL student learning new vocabulary [insert terms]. Create one short passage (150 to 200 words) that incorporates these words in the context of [theme of the unit].”
• Step 2. Upload the passage to the Microsoft Teams for Education’s learning accelerator (Reading Progress). Students record themselves reading the AI-generated passage aloud and respond to reading comprehension questions.
• Step 3. A feedback report evaluating various aspects of their speaking performance is automatically generated, allowing you to analyse common speaking issues and tailor support.
• Step 4. Give the report to students and encourage them to reread the passage after reviewing the feedback. 

This method encourages participation and provides students with an opportunity to practise speaking English beyond the classroom. Constant and timely AI-assisted feedback practice helps students assess their fluency and accuracy level and make continuous improvements. This approach is particularly beneficial for students with a lower speaking level who might feel anxious about speaking in public because they can acknowledge their specific strengths and weaknesses with reference to the feedback given.

• Spotlight guide: Bringing GenAI into the university classroom
• Beyond Chat – how AI teaching assistants are transforming student support
• Three ways to develop students’ AI literacy

Idea generation

The generation and organisation of ideas for academic discussion can be overwhelming, especially when studying in a foreign language. To address this, I encourage the use of GenAI for brainstorming ideas, allowing students with emerging English proficiency to explore topics and generate ideas without the immediate pressure of articulating them in group discussions. Here’s how you can do the same:

• Step 1. At the beginning of a discussion, encourage students to analyse the topic and brainstorm ideas independently using mind maps or flow charts.
• Step 2. After initial idea generation, students can use GenAI to brainstorm additional ideas, using prompts such as, “I am a first-year EFL student preparing for an academic discussion on [insert topic]. Provide me with two to three ideas supporting and opposing the topic, along with explanations and examples.”
• Step 3. Once they have generated ideas from both methods, students compare their initial thoughts with the AI-generated suggestions, helping them refine and finalise their ideas before the discussion.

Less time spent formulating ideas means more time to focus on language refinement. This approach also encourages students to think from multiple angles and builds their confidence, helping them become less hesitant to speak.

By providing structured scaffolding and feedback on academic speaking, these AI-assisted teaching strategies enable students to develop their skills with greater confidence.

Ka Ho To is a language lecturer in the School of Languages at Xi’an Jiaotong-Liverpool University.

If you would like advice and insight from academics and university staff delivered direct to your inbox each week, sign up for the Campus newsletter.