Anara AI’s Ads Are Undermining What a PhD Is Supposed to Teach You

AI can support research, but it should not be used to replace the core academic skills grad students are supposed to develop.

Let me start out by saying that I’m not anti-AI. Not at all. But I believe AI should be used as a tool to help with things we already know how to do. Not as a process for bypassing the learning stage.

The AI tool that prompted this post is Anara (which I am linking, but most definitely not advertising), a new AI tool for summarizing and citing research. The tool allows you to upload multiple research articles and then use a chatbot to ask questions about them. It has an autocomplete function that finishes sentences for you. It can paraphrase. It provides a map of how papers are connected. It can even provide a list of suggested citations for what you’ve written, where you can click a button and automatically cite from its list without ever having read the paper. (No, that’s not how citing works).

My specific concern with Anara is how it is being marketed to PhD students. I see ads for Anara on my TikTok and Instagram feeds daily. The company has partnered with many PhD/academia-focused creators, and they’re selling it quite convincingly as a tool that can help PhD students write their papers and literature reviews… but with, in my opinion, the wrong message. Anara’s ads overwhelmingly appear to be saying: Don’t bother reading papers yourself; just upload them, and the AI will do the thinking for you!

AI in research isn’t inherently bad, but it is when it’s being used instead of developing core skills, such as how to read, critically analyze, synthesize, and cite literature. Using shortcuts for these skills instead of developing them undermines the entire purpose of graduate education and is bad academic practice.

If you’re doing a PhD, I assume your goal is to become an expert in your field. Seeking the fastest answer by outsourcing thinking has never produced an expert. You become an expert by learning how to read deeply, spot methodological flaws, weigh evidence, and build arguments.

Perhaps your intelligence got you into your program, but it won’t carry you throughout your career.

You need to develop skills in asking the right questions, seeing connections in complex situations, making sense of contrasting conclusions, identifying limitations… These are not inborn talents. They are skills that need to be learned and developed. Having AI do them for you might allow you to produce a paper that convinces others that you have these skills, but it won’t actually give you the skills. And you’ll need them. Where’s that AI going to be when you’re listening to talks at a conference or having a real-world conversation about your field?

In one particular video about Anara (that I unfortunately had to watch for research purposes), the creator shared a list of questions that PhD students could ask the tool about their papers, including:

• What are the main findings?

• What are the common methodologies?

• What are the contrasting results and differing conclusions?

• What are the limitations and criticisms?

• What would be a good area for improvement?

All great questions to ask when you’re reading a paper.

Also the exact kinds of questions PhD students are supposed to learn to ask and answer themselves.

For someone who already has these skills and is in a time crunch for some reason, I can see how this AI tool could be useful (although, having read tens of thousands of academic articles by this point, I don’t think it would be faster for me—in fact, it would probably take me longer as I’d need to fact-check everything it gave me). But relying on tools like this too soon, before you’ve developed the skills to answer these questions independently, doesn’t provide any real efficiency gains. All it does is prevent the development of foundational skills.

If you’re skipping the part of grad school where you develop the analytical capacity that will shape your research, career, and credibility… why are you even there? (No, seriously. Why?)

The literature reviews you write in grad school help teach you how to do research. They’re not assigned as busywork but as a way to help you learn how to think as a scientist, sort through complex information, identify gaps, develop arguments, situate your work in a broader context…

Automating these tasks with a $12/month AI subscription is like paying someone to go to the gym for you and still expecting to get abs. Your intellectual muscles build through repetition. Use them.

There’s ethical issues here, too. Anara lets you upload any PDF, regardless of copyright. In doing so, you're potentially training an AI model on someone else’s work without their permission. That might be common these days, but it's still ethically murky. And unacceptable, in many cases. Would you be OK with someone uploading your unpublished or paywalled article into an AI tool that can learn from it and repackage it without your consent? A lot of authors would say no. Something to think about.

Another Anara ad demonstrated someone inserting references into their writing without reading the papers, stating that you could just choose a citation that you like.

That borders on misconduct (actually, it might just be outright misconduct).

If you haven’t read a paper, you can’t vouch for what it says, you can’t know whether it truly supports your argument, and you don’t know if you agree with the author’s conclusions.

Citing things you haven’t read and summarizing papers you haven’t analyzed also introduces massive risks of misinterpretation, misinformation, and bias. Even if the tool is accurate (remember, AI tools still hallucinate), citing this way encourages cherry-picking. Writing something and then finding a source that backs it up immediately brings bias into your work because you’re specifically seeking sources that agree with you.

I’m not saying there’s no place for tools like Anara. Used responsibly—after you’ve read the paper, after you’ve developed your analysis—they could potentially help with organization, synthesis, even idea generation. They could help you manage information overload once you’ve built the skills to know what to do with it.

But that’s not how Anara is being marketed.

It’s being marketed as a substitute for reading, writing, and analyzing, sending the message that critical thinking is optional, intellectual labor is outdated, and success comes from speed, not substance.

Academia is already hurtling toward a giant trash fire. We don’t need to pour lighter fluid on it.