From ultra-fast briefings to consulting 'dolphins': 6 Big Four leaders share how they use AI at work [Business Insider]
When the leadership of the Big Four accounting firms—Deloitte, PwC, EY, and KPMG—starts talking shop, you expect jargon about compliance and risk matrices. But over the last six months, I’ve sat down (virtually) with six of their top partners. The conversation has shifted. Now, they’re talking about “dolphins,” ultra-fast briefing bots, and why the best use of AI is often just getting your inbox down to zero.
These are not theoretical PowerPoint slides. These are real, daily workflows. Here are six specific ways these leaders are using AI right now—and what they say you should steal for your own job.
1. The “Ultra-Fast Briefing” for the Time-Crunched Partner
One partner at Deloitte told me he used to spend 90 minutes every morning reading analyst reports, industry news, and client updates. Now, he does it in seven. He built a simple internal bot that scrapes his personalized news feed, summarizes earnings calls, and even flags regulatory changes relevant to his specific three clients. “I don’t read the report anymore,” he said. “I read the bot’s summary. If I need to challenge the summary, I click the source. It saves me about 15 hours a month.” The key? He trained the bot to ignore fluff. No generic business news. Only actionable data.
2. The “Dolphin” Approach to Brainstorming
One of the most fascinating conversations was with a PwC advisory lead who uses AI not as a calculator, but as a creative partner. He calls it “consulting the dolphin.” The origin story? He was stuck on a client problem involving supply chain disruption. He fed the problem into an LLM (a large language model) but gave it a weird constraint: “Think like a marine biologist who knows nothing about logistics but understands swarm behavior.” The AI spit out an analogy about dolphin pod movement. It wasn’t the answer—but it unlocked a new way of thinking about distributed decision-making for his client. “I don’t use AI for the right answer,” he said. “I use it for the angle I hadn’t considered. It’s my cheap brainstorming partner.”
3. The Email Zero Machine (with a Human Filter)
KPMG’s head of innovation shared a brutally honest confession: “I hated my inbox. I had 300 unread emails by Tuesday morning.” His fix wasn’t an auto-reply. It was a custom AI filter that prioritizes emails by urgency and sender relationship. The AI drafts short replies for the bottom 80% of emails—internal approvals, status updates, logistics. He reviews them in a batch at 6 PM. “I don’t let it send anything without my click,” he stressed. “But it cuts my drafting time by 70%. I now spend that time actually talking to my team.” The rule: AI handles the routine; humans handle the relationship.
4. The “Red Team” Auditor for Contract Review
One EY partner in risk advisory described a tool that acts like a paranoid junior associate. When reviewing a vendor contract, the AI doesn’t just read it. It actively attacks it. It looks for hidden auto-renewal clauses, ambiguous liability caps, and terms that conflict with existing agreements. “It’s like having a lawyer who has read every contract we’ve ever signed in the past five years,” she said. “It finds anomalies I would have missed on a Friday afternoon.” The AI flags the risks; the human decides whether to fight. She estimates it has cut contract review time by half while catching errors that previously slipped through.
5. The Personal “Career Coach” Bot
This one surprised me. A Deloitte managing director uses a private, off-the-record AI journal. Each week, he inputs his calendar, his feedback from peers, and his own reflections. The AI analyzes patterns: “You had three high-stress meetings this week. Your feedback score was higher when you prepared a one-pager in advance.” It also suggests professional development resources—specific books, courses, or people to meet. “It’s not a substitute for a real mentor,” he said. “But it’s like having a personal data analyst for my own career. It reminds me of the stuff I’m too busy to notice.”
6. The Live “Translation” for Global Teams
Finally, a KPMG partner who runs a team spanning India, the UK, and Brazil uses AI for real-time meeting translation—but with a twist. The AI doesn’t just translate words. It translates tone. If a colleague in São Paulo speaks in a more direct, informal style, the AI flags it for the UK team: “This may sound blunt, but it is culturally normal here. No disrespect intended.” The partner says it has cut down on cross-cultural misunderstandings by a significant margin. “We used to have passive-aggressive email chains about tone,” he laughed. “Now, the AI acts as a cultural diplomat. It just makes the work less exhausting.”
The Common Thread: AI as a Tool, Not a Boss
If you read these six examples closely, you’ll notice a pattern. None of these leaders are handing over decision-making. They are using AI to reduce noise, find unexpected angles, handle grunt work, and smooth human interactions. They still make the calls. They still have the tough conversations. They still own the relationship with the client.
The AI is the junior analyst, the brainstorming buddy, the cultural translator, and the email secretary. But the human is still the leader. That’s the only way this works—when you treat AI like a very fast, very obedient, occasionally weird intern, not a replacement for judgment.
What’s the one thing you can do tomorrow? Pick one task you hate or one pattern you keep missing. Build a tiny bot. Give it one job. See if it makes your inbox a little emptier or your thinking a little wider. That’s all it takes to start.
Ahmed Abed – News journalist