I tried Claude Cowork in my Gmail inbox after Gemini got choked – and it saved me hours of work

Follow ZDNET: Add us as a favorite resource on Google.

Highlights taken by ZDNET

• Claude Cowork turned inbox chaos into actionable article research.
• Gmail search struggles with context and insight.
• Human verification is always important before using AI-derived quotes.

How many times have you had to go through a bunch of emails to find the messages you need? I’m not talking about searching for one or two by a specific word. I’m talking about getting a set of themed email messages from the thousands in your inbox.

I should have done this a few weeks ago. Because of my role as a tech journalist, my inbox is overflowing with people and companies pitching me on something or the other. Because part of my rhythm is AI, my Gmail gets criticized whenever I write about something a company executive or PR lawyer wants to be included in. And, you know what? Sometimes I find those places very important.

Also: Anthropic’s Claude Cowork goes to the cloud as data shows 90% of sessions aren’t coding

Take, for example, the day last month when it emerged that Anthropic was pushing Fable 5 users down to Opus. A number of people, including real experts on the subject, sent me emails with their comments about the disclosure. That was a busy week. I got over 7,000 emails, almost all of them related to something related to AI, Claude, Fable, etc.

Gemini for Gmail has limitations

In the past, I was spending hours sifting through my emails to find the information I needed. But this is the age of AI, so I asked Gmail to do it for me. Sometimes, Gemini for Gmail does a great job.

Not at this time. This time, it failed spectacularly. Here’s what I asked you:

Look in my Gmail promotions tab. I received a lot of emails from people who have opinions or want to be quoted about Anthropic's Fable 5. Please summarize the basic pitches for each of them (going back to Monday). Highlight any quotes or pitches that push back against Fable's restrictions as either too restrictive or not restrictive enough.

I think Gmail just couldn’t understand the context in each message. It does well when I ask it to receive simple messages or topics, or the latest email from the company. But this involved a request for understanding.

Gmail was not yet up to the task. But Claude Cowork was. Here’s how that works.

It allows Claude to be free from my email

This was the really hard part. Emotionally, of course. If there’s anything I consider my secret crown jewel, it’s my email archive. I’m comfortable with letting Gemini get away with it. I think since it works on Google’s infrastructure, Google has probably already integrated anything juicy into my Gmail. That ship has sailed.

But letting Claude send my email took a leap of faith. I’m not good at leaps of faith. So this was difficult. But my email has also been a bottleneck. If Cowork can save me half a day of tedious mail sorting, I might be worth it.

Also: I dropped Anthropic’s Claude Cowork into my files, and it was brilliant and terrifying.

What was ol’ Ben Franklin said? “Those who would sacrifice essential Liberty, to purchase temporary security, deserve neither liberty nor security.” It’s still a long way off, but letting Claude loose on my email will free up some of my time, and Anthropic’s reputation for security seems to indicate that the process will be pretty safe. Hopefully, I don’t offer freedom or safety.

To make this work, you’ll need to connect Claude to Gmail using a connector. Go to Claude’s settings, then Contacts, and select Gmail. As you can see, I gave Claude permission to read my Gmail, but not write (send) anything.

To keep my paranoia in check, I’ve turned off the connector once this project is complete.

Putting Cowork to Work

My challenge was twofold: first, make sure Cowork was generating the value I needed. Second, don’t get so caught up in making Cowork do what I want that I spend more time playing with it than I would if I just assembled what I needed by hand. Yes, the trap I fell into so many times.

I am an engineer. It is a real hazard at work.

I used the same information I showed above. In just over a minute, Cowork identified 12 areas of PR. Cowork also noted, “many of these are cybersecurity vendors who use the launch as a news hook.” Do you think?

Also: I sent Gemini and Claude to write my email responses – but only one sounded like me

The good news is that most people who try to use a news event as a starting point to achieve their goals usually include a useful idea related to the event they are starting. That’s what makes them powerful sources of media coverage.

My first exposure to this condition was back in one of my first jobs. My boss, the product manager, was always betting on our PR person to get “the word.” Basically, he wanted to get his name in the press as often as possible. Sure, the company name and products were important, but his goal was personal promotion. If he could be quoted or mentioned, he would be better known.

This is all part of the strange world of PR. Companies pay PR firms out of their company budget. But executives at those companies often try to leverage their companies’ contributions and use related PR efforts to boost their own careers. And we writers try to help managers and companies get more value for our articles.

It works because the more ideas we can present about a news event, the more circular the coverage becomes. In this case, I specifically asked Cowork to highlight emails that backfire against Fable’s restrictions as either too restrictive or not restrictive enough.

Also: I used Gmail’s AI tool to do hours of work for me in 10 minutes – with 3 commands

My next request was for Cowork to filter those relevant sites and identify those that were informative enough for me to use directly in my article. But I also wanted to make sure I had permission to publish, so here is the following information.

Find the pieces that contain full, on-record statements with explicit permission to publish, especially emails that say "free to use in any piece" or "free to use in coverage."

This gave me a set of email messages that might be useful to cross-reference, and message content that I knew I had permission to use.

Next, I wanted Cowork to pull good presentations and give me a clear list. But the key was that I didn’t want it to change people’s quotes. Don’t worry. “Trust, but verify” is very applicable here. I have built a great deal of information into several sections.

First, I set up the loop. It tells Cowork that it should generate a list of entries and do so from the email messages it previously verified.

For each validated email message, use this format. Separate each of the quotes with two blank lines. Do not apply any styling to any of the formatting.

Next is the section that tells Cowork how to format the personal information at the beginning of the entry. I like this method because I can quickly scan and determine if a person’s title or company makes that person approve the article.

The general format will be one of the following:

Name of person
Title of person
Company of person
Note: Be careful. This sort of email is often sent by a PR representative who is a different individual from the person being considered for a quote. Be sure you're listing the company executive or the subject matter expert, not the public relations contact.

The following is a request to extract the meat of the phrase. This is the information I have considered working in this article.

Extract a usuable statement, word-for-word, no changes. If any of the quotes are particularly relevant to usage concerns about Fable and need a second paragraph, put it in a second paragraph.

Notice I quoted both verbatim and no changes. Will it match this? Who knows?

Follow that with a URL to the company or, if you can't find one, a URL to the person's LinkedIn account.
Follow that, on its own line, with a link to the original email address
Follow that with a link to the email message itself

That last line is key. I wanted AI to sift through all my emails and extract actionable statements, but it’s still my responsibility if a quote fits into the subject. So when I found a quote on the Cowork list that I wanted to use, I used the link provided to open the email message and manually verify all the information before I used it.

This is very important. I used Cowork as a research assistant, not as a writer.

It’s worth it!

I have to say that this was very valuable and definitely a tool that I will use again. My partner did all the complaining in a thankless and boring job that often took me hours and hours. It sifted through thousands of email messages and provided me with a list of eight verified sources and citations.

If you want to see the article where I used this, here it is: Claude Fable 5 secretly threw AI researchers, and the internet went wild

The time it took me to carefully go through the list of eight sources was only a few minutes. I could devote most of my time to writing, with the bonus of not turning up and being depressed because I had just spent most of the day doing glorified clerical work.

Also: I tested ChatGPT vs. Claude to see which is better – and if it’s worth changing

Anthropic calls this “work around work,” which I think, as a concept, has legs. I will call Cowork again if I have some heavy admin work that I want to outsource.

Would you let Claude Cowork read your Gmail if it could save you hours of tedious research? Let us know in the comments below.

You can follow my daily project updates on social media. Be sure to sign up for my weekly update newsletter, and follow me on Twitter/X at @DavidGewirtzon Facebook at Facebook.com/DavidGewirtz, on Instagram at Instagram.com/DavidGewirtz, on Bluesky at @DavidGewirtz.com, and on YouTube at YouTube.com/DavidGewirtzTV.