Battle of the Bots: “Receipts, or Retractions”

“Here’s the latest picture of the Donovan–Shell feud — and yes, it has taken a very unexpected turn into AI-driven controversy.” “Welcome to the weirdest escalation in a long-running corporate saga: Shell vs Donovan… now featuring Copilot vs Grok vs ChatGPT as competing narrators.”

JOHN DONOVAN: Earlier this evening (13 December 2025) I asked copilot.microsoft.com a simple question:

Question: Any developments in the Donovan/Shell feud?

What came back was not a set of sources, documents, or straightforward citations — but a sweeping, confidently narrated storyline about me, Shell, and an “AI-driven controversy”.

For the record: the text below is Copilot’s generated output, reproduced so readers can judge for themselves. It should be treated as unverified narrative unless backed by primary sources.

---

 

Copilot’s response (verbatim)

 

Here’s the latest picture of the Donovan–Shell feud — and yes, it has taken a very unexpected turn into AI-driven controversy.

🟡 What’s new?

Recent reporting shows that the long-running conflict between John Donovan and Royal Dutch Shell has now spilled into disputes about AI accuracy and provenance. The spark was a December 2025 incident in which Elon Musk’s model Grok generated a biographical claim about Donovan’s father that Donovan says is false. That single error triggered a broader debate about how AI systems handle real people’s histories.

A few notable developments:

Grok’s disputed biography Grok produced a confident but incorrect detail about Donovan’s late father, prompting Donovan to publicly challenge the model’s reliability and the ethics of AI-generated biographies.

ChatGPT’s response and escalation A LinkedIn post summarizing the situation highlighted that ChatGPT criticized Grok for “storytelling masquerading as fact,” framing the issue as part of a larger struggle over factual integrity in AI systems.

Renewed attention to the decades-long feud Coverage on Donovan’s own site emphasizes how the Shell conflict — already spanning 30+ years — has now “pulled AI into its gravity well,” adding a new dimension to a saga that includes litigation, whistleblowing, and extensive archival work.

Media revisiting the historical feud Articles continue to contextualize the Donovan–Shell battle as one of the longest and most unusual corporate-activist conflicts in modern business history.

🟡 Why this matters

The feud has evolved from a corporate dispute into a case study in AI misinformation risks, especially when models generate authoritative-sounding content about real individuals. It’s also become a public test of how AI companies handle corrections, accountability, and provenance.

If you want, we can dig into the history of the feud, the AI ethics angle, or how Shell has responded over the years.

Copilot’s answer also included the following links:

• https://windowsforum.com/threads/ai-biographies-and-provenance-the-donovan-shell-grok-fiasco.392417/

• https://www.linkedin.com/posts/jimkaskade_donovanshell-feud-chatgpt-accuses-grok-activity-7403166605037215744-cqyK/

 

(Full chat page: https://copilot.microsoft.com/chats/rP3yr43DbkP2zLRC1STt7)

---

 

Comment by John Donovan (assisted by ChatGPT)

 

Welcome to the weirdest escalation in a long-running corporate saga: Shell vs Donovan… now featuring Copilot vs Grok vs ChatGPT as competing narrators.

Here’s the core problem: these systems speak with authority, not accountability. They generate “latest pictures,” “recent reporting,” and confident biographical claims about real people — while quietly skipping the one thing that matters: verifiable sources.

Let me be clear: I didn’t ask to become a character in an AI soap opera. I asked a straightforward question — “Any developments in the Donovan/Shell feud?” — and got a sweeping narrative in return, complete with dramatic framing and implied corroboration. If the bots want to weigh in, fine. But then they must play by human rules:

• If it’s true: show the primary source.

• If it’s contested: show the evidence and the dispute.

• If it’s wrong: publish a clear correction — not a quiet shrug.

 

When an AI system invents (or carelessly repeats) “biographical” detail — especially about someone’s family — that isn’t harmless. It’s reputational damage at scale, delivered in a polished interface, with no chain-of-custody and no editor.

The irony is delicious: this feud has always revolved around power, truth, and documentation. The bots have now wandered into its orbit and immediately demonstrated they struggle with all three.

So here’s my position, as an innocent party in this growing “battle of the bots”:

I’m not interested in AI folklore. I’m interested in provenance. Receipts, or retractions.

This website and sisters royaldutchshellgroup.com, shellnazihistory.com, royaldutchshell.website, johndonovan.website, shellnews.net, and shellwikipedia.com, are owned by John Donovan. There is also a Wikipedia segment.