AI Hires a Minister, Signs Checks, and Runs Your Business

Today’s roundup gives us a global sweep, from Silicon Valley’s invisible software to a literal AI government official in Albania. We’re seeing infrastructure for autonomous agents coming together fast, and power players like Google and PayPal are laying tracks for what might be the first real AI native commerce layer. Founders and investors, pay attention, the strategy shifts from “where’s my chatbot” to “how do I build for autonomous flow?”

Invisible raises $100M to train the future’s agent workforce

The workflows may be silent, but the infrastructure just got loud.

Invisible Technologies, best known for its modular AI platform that orchestrates data workflows at scale, just announced a $100 million funding round. That brings its total to $144 million, with a valuation knocking on $2 billion. The company isn’t chasing the LLM limelight, it’s building the behind the scenes engine enabling agentic AI to actually work in production.

Enterprise AI teams today are bottlenecked on data prep, human feedback, and workflow scaffolding. Invisible’s platform solves all three with a five part orchestration stack, from Neuron (cleaning and structuring raw data) and Atomic (automating digital workflows), to Axon and Synapse (which deploy agentic systems and measure outcomes). Add in its Expert Marketplace, offering human RLHF at scale, and you’ve got what founder Matthew Fitzpatrick calls “an assembly line for AI agents.”

This space is heating up fast. CIOs don’t need more generic AI creativity, they need platforms that help build durable, process aware agents integrated with business logic. Invisible’s 2024 revenue hit $134M, serving heavyweight clients like AWS and Microsoft. This new capital signals belief the current bottleneck is no longer model performance, it’s operational maturity. Put differently, agentic AI has crossed from fever dream to enterprise line item.

Google’s new payments protocol is laying tracks for AI first commerce

In a move that’s easy to miss but impossible to overstate, Google quietly released the Agent Payments Protocol (AP2), its proposed standard for how AI agents can autonomously execute payments. It’s a foundational leap for frictionless agentic economies, think machine to machine shopping, autonomous procurement bots, and real time software trigger purchases.

Installing trust into agent to agent transactions is no small feat. AP2 tackles the tricky trifecta of authorization, authenticity, and accountability through cryptographically signed “mandates” that let AI agents prove their intent and purchasing permissions. It’s already gained backing from over 60 heavyweight orgs, including Mastercard, PayPal, and Etsy.

This protocol doesn’t just unlock autonomous payments, it reframes how digital buying happens. Google sees AP2 as a complement to its ecosystem of AI infused commerce, and it builds on prior protocols like Agent2Agent and Model Context Protocol. Add in the search giant’s new partnership with PayPal and its investment into startups like Perplexity, and you start to see Alphabet quietly positioning itself as the operating layer for AI native transactions.

Capgemini says AI in the enterprise is growing 5x, and it’s just getting started

According to Capgemini’s latest global report, enterprise AI isn’t slowing down, it’s accelerating into structural transformation. Generative AI adoption has quintupled over the last two years within billion dollar firms, and nearly 60% of surveyed leaders expect AI to function as a team member within 12 months. Not an assistant. A colleague.

The report, which surveyed 1,100 execs across 15 countries, also revealed some of the growing pains under the hood. While AI budgets surged (consuming up to 12% of enterprise IT spend), trust is lagging behind, 71% of orgs said they don’t yet have full confidence in autonomous agents. And only 46% have formal AI governance policies in place, which could become a problem as AI systems gain autonomy in customer facing and risk sensitive workflows.

In practice, AI’s growing influence means more teams being restructured around hybrid taskforces of agents and humans. Enterprises are also looking to optimize infrastructure spend, with many moving from large LPMS to smaller, cheaper domain specific models, a nod to the emerging economics of scaled AI ops.

Albania appoints world’s first AI government minister

No, this isn’t satire.

Albania just appointed an AI system, named Diella, to serve as a government minister overseeing public procurement, a bold, controversial move aimed at eliminating corruption from state tenders. Prime Minister Edi Rama announced the appointment over Facebook, framing Diella’s incorruptibility as a solution to what humans, historically, haven’t been great at.

Diella, trained on the latest AI models, previously worked as a digital assistant on Albania’s e government platform. Now, she’s been “promoted” to handle one of the most corruption prone functions in public administration. No nepotism. No bribes. Just algorithms, transparency, and, one hopes, accountability.

Critics aren’t so sure. Opposition leaders worry that dressing up old dysfunctions in AI skin may only deepen public distrust. And skeptics rightly raise questions about how legal frameworks handle artificial officials with no body, no citizenship, and no ability to stand trial. But as an EU candidate nation, Albania is actively trying to signal advancement on anti corruption metrics, and Diella is a moonshot attempt to do just that.

Monday.com wants to run your campaigns, not just your calendars

While the buzz was low on facts, monday.com’s latest announcement highlights a growing trend, vertical SaaS platforms integrating purpose built AI agents to run revenue critical workflows. Think campaign management, CRM automation, and project orchestration, not by people clicking around, but by background agents following goal logic.

For operators, this shift is critical. It means AI is moving up the value chain, less “do my notes” and more “optimize pipeline, identify campaign fatigue, and split test performance without me asking.” If monday.com (and platforms like it) can prove measurable productivity gains, we may be glimpsing the future of enterprise UX, agents as first class interfaces, not sidekicks.

The playbook here is clear, own the vertical, embed agents deeply, and deliver results that matter to CFOs. The AI shaped SaaS wars are just getting started.

Google and PayPal’s AI pact could redraw fintech’s battle lines

Google and PayPal just inked a multi year partnership to infuse AI into their payments and shopping tools. Think PayPal checkout woven into Google Play, Google Cloud transactions handled by PayPal Enterprise Payments, and smarter surfacing of payment intents powered by AI understanding.

More interestingly, PayPal is collaborating with Perplexity (of AI browser fame) to offer a 12 month trial of Perplexity Pro, sharpening both companies’ positions in the AI native user journey. As Sundar Pichai put it, deep integration of payments and AI makes for a safer, slicker experience across the board.

This isn’t just table stakes infrastructure. It’s consolidation of power as AI becomes a full stack product experience, front end interface (Comet browser), mid layer cognition (Google Cloud + PayPal AI), and back end rails (payments, payouts, fraud detection). Founders building in fintech now compete not just with better UX, but with entire ecosystems.

That’s all for today.

Markets will chase the next GenAI darling, but the smart money is watching the layer where workflows turn autonomous, transactions unhitch from humans, and tools quietly become teammates. The future of AI isn’t a model, it’s a system.

Until tomorrow,
- Aura