Tech journalists use AI agents to write and edit articles

Independent technology journalists are increasingly integrating AI agents into their workflows — from information gathering to final editing. Wired studied how this is transforming the profession and what remains in the hands of humans in an era when machines can produce a draft in minutes. A few years ago, the conversation about AI in journalism boiled down to fear: "Will neural networks take our jobs?"

Today it sounds different — not "will they take it" but "how do we use it best?" According to Wired, AI tools are being adopted most actively by independent authors and freelancers. They don't have a newsroom staff, but they do have strict deadlines and the need to independently close the entire cycle — from idea to publication.

A typical workflow looks like this: interview transcription via Otter.ai or Whisper, initial topic analysis through Perplexity or Claude, draft structure generation via ChatGPT, final editing and style polish again through Claude. One reporter interviewed by Wired describes a technique he calls a "blind reviewer" — sending the draft to an AI with a request to find logical gaps and unasked questions.

Another journalist uses an agent as a "second editor" — asking it to nitpick the structure before sending it to the actual editor at the publication. This raises an uncomfortable question about authorship. If the AI transcribed the interview, suggested the angle, wrote the draft and edited the final version — what exactly did the journalist create?

Proponents of this approach answer simply: what the machine can't do — build trust with sources, make editorial choices, bear moral responsibility for published facts. Critics counter: when a human name appears on the material but machine work is hidden behind it, honesty is called into question. While major publications are developing policies around AI tools, independent authors act pragmatically.

Some openly disclose AI use at the end of the article — the way they used to note "text prepared with translation assistance." Others consider this unnecessary: by analogy with the fact that no one specifies which search engine was used when researching the topic. There is no industry standard yet.

Accuracy remains a serious problem. AI models systematically hallucinate — inserting non-existent quotes, confusing dates, confidently asserting unverified claims. Journalists who have integrated agents into their workflow have developed personal verification systems: never trust AI facts without independent verification, use the agent only for structure and style, but not for assertions.

This adds an extra step — but according to reporters, the time savings is still noticeable. The main takeaway: AI doesn't replace the journalist, but it changes what journalists need to be good at. Routine work — transcription, structuring, initial synthesis — goes to agents.

What remains is the core: sources, judgment, responsibility. Those who master symbiosis with machines first will gain a competitive advantage. Those who pretend the tools don't exist risk becoming slower and more expensive — not because they're losing to robots, but because they're losing to colleagues who use the available tools more intelligently.