PLUS: AI assistants now buy for you, and your new coworker is now AI too.

Morning All, Apple and Google's surprise partnership reshapes the AI landscape - with Gemini now set to be the engine that powers Siri in the future. Long awaited and long overdue, but still very welcome...fingers crossed, Siri may not suck for very much longer.

This deal suggests even tech giants need partners to compete in today's AI race. It also gives more evidence to something I've been thinking for a while...ChatGPT might get all the hype and publicity but the real value lies in the products and services that will be transformed by GenAI.

Today's dots:

• Apple's Gemini-powered Siri overhaul
• AI shopping adoption reaches critical mass
• Claude's Cowork assistant bridges productivity apps
• Finance chiefs accelerate AI deployments
• Microsoft's community-first data center pledges

Siri’s Brain Just Got a $4T Upgrade

Here's the thing: Apple just bet big on Google's Gemini AI to rebuild Siri in a landmark partnership that merges Apple's hardware reach with Google's conversational tech - officially confirmed days before its delayed 2026 rollout.

Let's unpack that:

• The multi-year deal makes Gemini Apple's foundational AI engine, powering everything from smarter Siri interactions to Apple Intelligence features, while maintaining Apple's privacy-first approach
• Google reportedly secured $1B/year licensing fees in the agreement - a major departure from Apple’s historical reluctance to partner with rivals
• The delayed Siri overhaul finally arrives late 2026, nearly 2 years behind competitors - with Bloomberg having first reported plans for a 1.2 trillion parameter Gemini model last November
• Apple confirmed all processing will happen either on-device or through its Private Cloud Compute infrastructure, maintaining its privacy safeguards despite outsourcing AI smarts
• The move sidelines OpenAI's ChatGPT integration as secondary functionality - though Apple have stated their existing partnership is still active

If you remember nothing else: This deal proves even Apple can't crack advanced AI alone in 2026. Google's infrastructure advantage just became Siri's backbone across 2.2 billion active Apple devices. This may only be the start but at the end of the race I'm betting on Google standing with their hands race...victorious.

AI Shopping Agents Hit Mainstream Adoption

Here's the thing: AI shopping tools just hit a critical tipping point with 41% of consumers now regularly using AI agents for purchases - and retailers like Nike are slashing inventory waste by 23% through real-time forecasting. (Source)

Let's unpack that:

• Adoption accelerated after ChatGPT, Gemini and Copilot rolled out one-click checkout last year - driving 700% YoY traffic growth to retail sites during holiday seasons
• Brands like Ridge Wallet now optimise product descriptions for conversational queries (e.g. "gifts for hikers") rather than keyword-stuffing, resulting in more natural discovery pathways
• Early adopters see 31% higher conversions from AI referrals vs traditional channels - justifying platforms' 4% transaction fees for direct integrations
• Shopify's Gemini integration lets 1M+ merchants tap into AI-native shoppers who prefer prompts like "find a sweltering wedding guest dress" over brand searches
• Inventory win: Predictive AI agents helped direct-to-consumer brands reduce overstock waste by nearly a quarter through better demand modelling

If you remember nothing else: Conversational commerce isn't coming - it's already dictating purchase journeys. Brands rewriting their content strategies for prompt-based discovery are converting and monetising shoppers without an "Add to Cart" button in sight.

Claude's Cowork launches: Your new AI office assistant

Here's the thing: Anthropic just released Cowork - an AI assistant that organises files, generates reports, and bridges apps like Chrome and Asana right from your Mac. Their research preview shows how this places Claude Code's agentic abilities in the hands of the masses.

Let's unpack that:

• Simply point Claude at any folder and watch it sort messy downloads, create expense sheets from screenshots, or draft reports from scattered notes - no coding required
• Unlike basic chatbots, it works autonomously like a human colleague - planning tasks, checking progress, and even handling parallel workflows
• Currently macOS-only for Claude Max subscribers, but includes crucial safeguards requiring approval before risky actions
• Pairs with Chrome for web-based tasks while optional 'skills' can improve document creation - perfect for SMBs drowning in paperwork
• Put this together with Google's new Universal Commerce Protocol and this is more evidence that 2026 is the year of AI agents.

If you remember nothing else: This shifts AI assistants from conversation partners to proactive coworkers that complete actual work. While early adopters should review safety guidelines, it dramatically lowers the automation barrier for non-technical teams.

Finance chiefs go all-in on AI deployment

Here's the thing: 87% of CFOs are moving AI from pilots to full deployment in 2026, with Deloitte's latest survey showing finance leading enterprise AI adoption.

Let's unpack that:

• Finance teams now prioritise digital transformation over risk management, deploying AI for fraud detection (37% faster than human teams) and forecasting accuracy improvements
• They're adopting hybrid quant architectures - blending large language models with classical algorithms like XGBoost for explainable financial predictions
• 50% plan internal promotions to upskill existing talent rather than layoffs, redeploying staff to AI oversight roles
• AI agents are being baked directly into workflows - automating 41% of invoice processing in early adopters
• Nearly all respondents see this shift making finance departments more strategic across procurement, pricing and M&A decisions

If you remember nothing else: Successful AI deployment isn't just about technology - it's about integrating human expertise with machine efficiency. This pivot transforms finance from cost-centre to strategic driver for entire organisations.

Microsoft flips the script on AI data centres with community-first pledges

Here's the thing: Microsoft just pledged to cover 100% of its AI data centre power costs after public backlash over rising electricity bills - a bold reset of tech infrastructure practices that could reshape how the industry balances innovation with local impact. Full announcement →

Let's unpack that:

• The Community-First AI Infrastructure plan commits to pay utilities directly for all operational electricity and grid upgrades - aiming to halt residential bill hikes seen in Virginia and Illinois where data centres expanded rapidly
• Water conservation gets a 40% efficiency target by 2030 including industry-first 'closed-loop liquid cooling' that completely eliminates drinking-water use in new facilities like the Washington state campus
• A new apprenticeships pact with North America’s Building Trades Unions will train local workers for thousands of construction and maintenance roles - while rejecting tax breaks to ensure schools/hospitals benefit
• Water repayment works like a bank account - Microsoft funds projects like leak repairs in Phoenix and wetland restoration in the Midwest to return more water than consumed. See how Quincy, WA benefits →
• Every major data centre location will get tailored AI education programmes via schools and libraries - building on successful tech bootcamps at Wisconsin’s Gateway Technical College

If you remember nothing else: This isn’t just goodwill - it’s smart crisis management as local opposition threatens £372bn AI infrastructure plans. Microsoft’s commitments could become the minimum expectation for tech’s physical expansion worldwide.

The Shortlist

Schneider launched LumenFI, an AI finance platform delivering real-time insights via generative AI and machine learning. Can CFOs finally trust automated financial intelligence?

Quantifind revealed 87% of financial crime units now prioritize 'explainable AI' models over raw accuracy. Is regulatory compliance becoming AI's killer app in banking?

Gradient Flow tracked hybrid quant architectures blending LLMs with classical algorithms like XGBoost for financial predictions. When will Wall Street fully embrace agentic workflows over human quants?