Google versus DuckDuckGo is an interesting battle. But for educators, the real question isn’t which search engine wins. It’s what students actually learn when they use them.
I’ll admit something. Google’s AI summaries often slow me down.
When I’m looking for reliable information, I usually want to read the original source. Instead, Google increasingly presents me with a confident-looking AI summary that I then need to verify.
So now I have two jobs: read the AI’s interpretation and check whether the original source actually supports it.
Sometimes that’s useful. Sometimes it’s just another obstacle between me and the information I wanted.
And that got me thinking about how we teach research in schools.
The Lazy Research Lesson
We’ve all seen it. Perhaps we’ve even taught it.
“Today, I’d like you to research renewable energy. Find five interesting facts and make a presentation.”
Students open Google, type in the topic, collect some information and paste it into slides.
Forty minutes later, everyone has a presentation. But what have they actually learnt?
The problem isn’t Google. The problem is the task.
A poorly designed research lesson can reward students for finding information rather than understanding it.
And AI makes that even easier.
Why spend twenty minutes gathering facts when an AI summary can deliver them in seconds?
AI hasn’t created the lazy research lesson. It’s just made it possible to finish one in five minutes.
Google AI Search: Helpful or a Hindrance?
Google’s AI Overviews can be genuinely useful. Ask a complicated question and you may receive a clear explanation that draws together information from several sources.
For someone unfamiliar with a topic, that’s a fantastic starting point.
But there are trade-offs.
The advantages:
- Quickly introduces unfamiliar topics and terminology.
- Summarises complex ideas in accessible language.
- Helps students formulate more specific questions.
- Can suggest connections that students might otherwise miss.
The disadvantages:
- Students may accept the summary without checking its accuracy.
- AI can remove uncertainty and disagreement from complicated topics.
- Students may never visit the original sources.
- A polished explanation can create an illusion of understanding.
- Checking AI-generated claims can take longer than finding a reliable source directly.
Research published in 2026 examining 11,000 search queries found that AI-mediated search systems can introduce biases in source selection and reduce expressions of uncertainty in summaries.
In other words, AI doesn’t just shorten information. It can change how that information is presented.
And sometimes the uncertainty is exactly what students need to understand.
DuckDuckGo: Back to Searching?
DuckDuckGo offers something increasingly unusual: a search experience where AI-generated answers can be switched off.
Its dedicated No-AI Search removes its AI search features and filters out AI-generated images as far as possible.
That means students are presented with search results rather than an AI-written synthesis.
The advantages:
- Encourages students to visit and examine original websites.
- Provides opportunities to compare multiple sources.
- Makes students practise identifying relevant information.
- Avoids placing an AI-generated interpretation between the student and the source.
- Gives schools and teachers more control over AI use.
The disadvantages:
- Finding good information can take longer.
- Students may struggle with unfamiliar terminology.
- Search rankings still influence which sources students encounter.
- Students can still copy and paste without understanding anything.
- A list of links doesn’t automatically develop critical thinking.
That last point is crucial.
Removing AI doesn’t magically make a research lesson good.
Students have been copying from Wikipedia, websites and search snippets for years. DuckDuckGo doesn’t solve that problem.
Good teaching does.
The Skills Are Different
Rather than asking which search engine is better, perhaps we should ask which skills we’re trying to develop.
| Traditional search | AI-assisted search |
|---|---|
| Constructing effective search queries | Asking precise, useful questions |
| Selecting relevant websites | Evaluating generated explanations |
| Comparing original sources | Checking AI claims against sources |
| Identifying bias and credibility | Recognising omissions and hallucinations |
| Synthesising evidence independently | Refining and challenging AI output |
Both approaches can develop valuable skills.
The danger is assuming that students are practising these skills simply because we’ve told them to research something.
Let’s Design Better Research Lessons
Instead of asking students to find five facts about renewable energy, what if we asked:
“Should our school install solar panels?”
Now students have a problem to solve.
They need to investigate costs, energy production, local weather, environmental impact and possible alternatives.
They must evaluate evidence, consider competing priorities and make a recommendation.
We could even divide the activity into three stages.
Stage 1: Search without AI
Students use traditional search to identify three relevant sources. They must explain why they trust each one and what information it provides.
Stage 2: Introduce AI
Students ask an AI tool to evaluate the same question. What did it identify that they missed? What did it overlook? Can its claims be verified?
Stage 3: Make a judgement
Students produce a recommendation supported by evidence, acknowledging uncertainties and limitations.
Suddenly, AI isn’t doing the thinking for them.
It’s something they have to think critically about.
And traditional search isn’t simply an exercise in collecting information.
It’s part of a genuine investigation.
The Real Problem Isn’t the Search Engine
If we ask students to find information, technology will keep making that easier.
Google made it easier to find websites. Wikipedia made it easier to access explanations. Generative AI makes it easier to produce answers.
We shouldn’t be surprised by any of this.
The challenge for educators is to design tasks where finding the answer isn’t the end of the learning.
Where students need to question, compare, evaluate, justify and make decisions.
A good research lesson isn’t about finding five facts. It’s about knowing which facts matter, whether they’re trustworthy and what to do with them.
And no search engine, with or without AI, can replace that.
Further Reading
- Google Search Help – AI Overviews – Google’s explanation of AI Overviews, their limitations and the Web filter.
- DuckDuckGo – No-AI Search – An alternative search experience without AI-generated answers.
- Answer Bubbles: Information Exposure in AI-Mediated Search (2026) – Research examining how AI search changes source selection, uncertainty and information presentation.
- Wired – Google Is Forcing AI Overviews in Classrooms – Coverage of the debate about AI search in schools.
What do you think? Has AI improved the way your students research, or has it simply made it easier for them to avoid the difficult thinking?
