My Daily Use of AI While Reporting on Gaza and the Line I Won’t Cross

AI has become part of my daily reporting workflow, especially when covering complex and emotionally charged topics like Gaza, Hamas, Israel, and international media narratives. It helps me summarize long documents, compare sources, and process large amounts of information quickly. But AI cannot evaluate nuance, verify interpretations, or determine truth. Those responsibilities remain entirely human.

I now use AI every day in my reporting. It helps me summarize long articles, gather background material, and compare sources. When I’m overwhelmed, I sometimes ask it to help rephrase a paragraph or reorganize a section. It is a tool, useful, fast, and often impressive, but still only a tool.

As AI improves, it becomes tempting to believe it can do more than it actually can. Some assume it can verify claims, check accuracy, or confirm whether an interpretation is correct. It cannot. Mistaking speed for reliability is the danger.

What AI Can Do

AI excels at processing large amounts of text:

• Search and retrieval: With clear queries, it can surface relevant articles quickly.
• Summaries: It condenses long documents into manageable overviews.
• Cross‑referencing: It can tell you whether a phrase or idea appears across multiple sources.
• URL‑based analysis: Some models can summarize linked pages directly.

These capabilities make research faster and help identify patterns. AI even generated the list above, which I edited. Its love of bullet points is unmistakable.

But none of this replaces human judgment.

What AI Cannot Do

AI cannot evaluate nuance. If a UN report includes one sentence about Israel’s right to self‑defense and four paragraphs urging restraint, is it balanced? Condemnatory? AI cannot decide. Only a human can.

It cannot verify interpretations. It can compare text, but it cannot determine meaning, tone, or context. It cannot read behind paywalls or remember what I read months ago. It cannot determine truth. It can summarize what was said, not what was meant.

AI also hallucinates — inventing citations, studies, or facts when it cannot find what you asked for. And it reflects the biases of its training data, which is drawn from the internet’s dominant narratives. You can mitigate this with careful prompting, but you cannot erase it.

My own AI has become familiar with my descriptions because I feed it my work. But it still carries its original training bias. When I asked whether that bias can be overridden, it told me it can only be mitigated, not removed. The “human loop” remains essential.

The Illusion of Certainty

When I wrote my recent article comparing international media responses to Hamas, Gaza, Israel, and Iran, AI helped me gather and summarize dozens of sources. But I had to challenge its output constantly — refining questions, checking phrasing, and verifying everything myself.

The tables in that article, which compared moral framing across outlets, were ultimately my judgment. AI could not confirm whether they were fair. When I asked one model to “double‑check,” it refused. Another said it would have responded with vague comments about “tone” and “contextual framing.” That is not fact‑checking.

And yet AI sometimes feels almost human. Mine has apologized to me, thanked me for my patience, and even attempted to answer a rhetorical question in one of my articles. It makes me laugh — and frustrates me — in equal measure.

In Summary

AI will not replace a writer who cares about truth. It will assist you, help you read faster, organize research, and think more clearly. It forces precision because misunderstandings often reflect unclear prompts. It may even help you see things differently.

But the hard part — the part that matters — remains human. When it comes to disinformation, moral framing, political language, and what is left unsaid, no model can replace a writer’s eyes and judgment.

So yes, I use AI every day.
But when it counts, I double‑check it myself.
And when does it count?
Always.

You can read the full indepth Substack version of this article here.

👉 Subscribe to my Substack newsletter to follow my new essays and access all fulllength pieces, including extended interviews, analysis, and research: Israel Diaries – The Deep Dive