• 420,000 employees billing hours while AI agents do the same work for $0.02/1000 tokens.
• Selling your US government unit after an ICE controversy — great timing, since AI was about to obsolete it anyway.
• Buying WNS for $3.3B in 2025 while OpenAI is giving away half your service catalog.
• 'Strategic agentic AI initiative with Google Cloud' — congratulations on discovering agents 18 months after everyone else.
• Your competitive advantage is that clients don't know how to use Claude Projects yet. Clock is ticking.

24-36 months until Fortune 500 clients realize they can replace $2M Capgemini engagements with a $200/mo Cursor subscription and an intern. The mid-market goes first.

• 'Co-develop AI-native digital solutions for manufacturing' — six words to say 'we're learning on your dime.'
• 'Revolutionize customer experience' — revolutionize is consultant-speak for 'implement the obvious thing competitors shipped last year.'

If Capgemini was a human, it'd be that friend who shows up to every party 45 minutes late with a PowerPoint about why they're actually early if you adjust for traffic.

• 420,000 employees = massive inertia buffer. You can lose 100K headcount over 3 years and still be huge.
• Enterprise relationships cemented in concrete — CIOs who've used you for 15 years won't switch instantly.
• Regulatory/compliance expertise in banking and healthcare buys 18-24 months while AI companies figure out GDPR.
• Brand equity: you're the safe choice. That's worth real dollars until it suddenly isn't.

What's the difference between a Capgemini consultant and an AI agent? The AI agent doesn't bill you for thinking time.

• Kill the bottom 40% of engagements by revenue — that's where AI-native boutiques will eat first. Go upmarket or die.
• Hire 5,000 ML engineers from FAANG in the next 12 months. Pay 40% above your current comp bands. This is existential.
• Ship an AI-native product (not services) in 6 months that clients can buy without talking to a salesperson. Prove you can build, not just integrate.