Mistral Lands €3B From Samsung, Europe’s Biggest Tech Raise Ever

AI Agents

Mistral Banks €3B From Samsung in Europe’s Biggest-Ever Tech Raise

What happened: Paris-based Mistral AI closed a €3 billion Series D round led by Samsung Electronics, at a post-money valuation above €21 billion — the largest equity raise ever completed by a European technology company. The EU-backed Scaleup Europe Fund (managed by EQT) and returning investor PSG Equity co-led, with ASML, NVIDIA, Bpifrance, BNP Paribas and BlackRock-managed funds also participating.

Why it matters: This is Europe explicitly buying itself a seat at the frontier-model table. Brussels has spent months warning about dependence on US and Chinese AI providers, and this round is the counter-move: a “sovereign AI champion” now funded to expand frontier research, add compute, and chase a one-gigawatt AI cloud footprint across the EU by 2030. For anyone building on or choosing between models, a well-capitalized European alternative changes the procurement conversation — especially in regulated industries and the public sector.

What’s next: Watch how the money splits between raw compute and actual model capability. Mistral’s “Forge” system (launched in March) already lets enterprises like Airbus, ASML and the European Space Agency train on proprietary data. A Samsung-led round also hints at deeper hardware-and-model bundling ahead. The real test isn’t the valuation — it’s whether the next frontier model can genuinely compete with US labs on benchmarks, not just on narrative.

Source: EUToday

Faye Pays Travel Claims Before You Even Ask — But AI Never Denies Them

Travel insurer Faye now detects flight disruptions automatically and pushes payouts to a digital wallet while you’re still mid-trip, aiming to fully automate over half its claims volume in 2026. The sharp design choice: its AI can approve claims but is never allowed to deny one — a human always makes the denial call, which CEO Elad Schaffer frames as a deliberate trust boundary, not a limitation. It’s a clean template for any business automating customer-facing decisions.

Source: PYMNTS

Your Agent’s Security Problem: 20 Years of Identity Controls Don’t Apply

Security practitioners warn that enterprises are handing AI agents privileged credentials, tools and network access while relying on controls built for humans — and agents act at machine speed, chain actions into unintended outcomes, and spawn sub-agents. The consensus fix: put the control outside the model (allowlists, egress proxies, short-lived tokens, human approval for high-risk actions) rather than trusting system prompts, because “if your security depends on a model choosing to behave, you have hope, not control.”

Source: CSO Online

AI News

Merchants Want AI’s Help — Just Not With Pricing

A PYMNTS/Visa survey of 1,185 merchants finds 46% rank final pricing as the function they’re least willing to hand to AI agents, with fraud, disputes and redirecting customers to rivals close behind. Meanwhile only 15% have product data structured well enough for an agent to read — so the near-term winners aren’t the flashiest agents, but the merchants who clean up their data first.

Source: PYMNTS

Logitech Ships a Physical Control Deck for AI-Assisted Coding

Logitech’s new MX Keypad gives developers customizable display keys that trigger prompt macros, refactor code and switch between AI tools — a hardware response to the “too many AI windows” problem. It’s a small but telling signal: as AI becomes a daily workflow layer, the interface is moving back toward physical, dedicated controls.

Source: Forbes

The AI Productivity Boom Is Landing in Employers’ Pockets, Not Workers’

A new report finds employers — not employees — are capturing most of the productivity gains from AI adoption, with the efficiency upside flowing to the bottom line rather than into higher wages. It’s the early shape of a debate that will only grow as agents take on more work: who actually owns the value AI creates.

Source: RTÉ