In partnership with

NEW LAUNCHES

The latest features, products & partnerships in AI

• Deutsche Telekom unveils Magenta AI search tool with Perplexity integration

• AI-powered confessional debuts at Swiss church

• Trimble launches major AI initiatives for construction and agriculture automation

• FlagEval is a new benchmark that assesses AI models’ ability to debate one another

• Google’s AlphaQubit addresses one of quantum computing’s biggest hurdles

• Wix unveils AI tool for agencies to create visual sitemaps

• Virgin Media O2 partners with Perplexity for AI search access

IMPLEMENTATION

Announcements, strategies & case studies

• New report details enterprise AI adoption trends and attitudes

• Forrester Wave Q4 2024: The trend toward AI-enabled, unified social marketing platforms

• Half of Gen AI users want the technology open-sourced — Here’s why

• 99% of firms plan to increase GenAI investments, study finds

• The new role of senior leaders in creating a data culture for the AI era

GOVERNMENT

Law, regulation, defense, pilot programs & politics

• AI, healthcare and NHS’ plan to implement digital patient passports

• US leads global AI ecosystem, Stanford study finds

• IEEE unveils new standard to assess AI system trustworthiness

• Anthropic CEO calls for mandatory safety testing on all AI models

• In Trump’s shadow: Nations convene in SF to tackle global AI safety

• AI discrimination lawsuit reaches $40M settlement

AI AGENTS

Launches, research & more from the AI Agent Report Newsletter

• Procore unveils AI agents for the construction industry

• Singapore researchers put Anthropic’s ‘Computer Use’ feature to the test

• Salesforce expands Agentforce with new AI testing tool

• xpander.ai’s new step-by-step system makes AI agent more reliable

• Satya Nadella: Microsoft’s AI agents are only in the ‘middle innings’

• AI agents: Game changer or overhyped tech trend?

• Orchestrator agents: The AI managing the next generation of AI

IN OTHER NEWS

Compelling stories beyond the usual categories

• AI company valuations are soaring despite signs of slowing progress

• FlagEval is a new benchmark that assesses AI models’ ability to debate one another

• AI PDF’s rapid user growth demonstrates the power of thoughtful ‘AI wrappers’

• Despite cooling issue reports, Nvidia says Blackwell AI chips are racing forward

• HBR: Can Gen AI destroy incumbent firms’ competitive advantage?

What’s happening in AI right now

Europe builds its AI strategy

Deutsche Telekom's integration of Perplexity AI marks a shift in how European telecom giants view artificial intelligence - not as a distant frontier, but as an immediate competitive necessity. When Spain's Telefonica also invested in Perplexity, it signaled that European industry is looking to integrate AI capabilities at scale

Infrastructure moves

Denmark's new Gefion supercomputer, with over 1,500 NVIDIA H100 GPUs, arrives as European nations push for sovereign AI computing capacity. Meanwhile, France's Mistral AI has launched a 124-billion parameter multimodal model to compete with American leaders. Their new Le Chat platform now includes web search, Canvas features, document analysis, and automated task agents - suggesting European firms can match U.S. capabilities with fewer parameters.

Public sector steps 

The UK's NHS is implementing AI-powered Patient Passports to digitize healthcare delivery at scale. The Vatican partnered with Microsoft to create an AI-powered digital twin of St. Peter's Basilica, processing over 22 petabytes of data from 400,000 high-resolution images. Both projects demonstrate how European public institutions are finding practical applications for AI.

Focused applications 

Greater Manchester's AI traffic cameras have already detected over 3,200 unsafe driving incidents in five weeks, including 812 drivers using mobile phones. The UK government's new AI assurance platform helps businesses evaluate and manage AI risks. Meta's smart glasses are rolling out across France, Italy, Ireland and Spain - though with more limited features than their U.S. counterparts due to regulatory considerations.

Search for alternatives

Ecosia and Qwant's collaboration to build a European web index challenges Google's search monopoly, with testing set to begin in France in early 2025. Mistral AI's rapid development shows Europe can build competitive foundation models. These efforts reflect a desire for technological capability without dependence on foreign platforms.

Industry progress 

In healthcare, Cambridge researchers are applying AI to develop better treatments for pediatric brain cancer, focusing on the roughly 420 children diagnosed annually in the UK. In finance, European fintech companies are reshaping credit scoring while adhering to strict consumer protection standards. The European Court of Justice has mandated human oversight in automated credit decisions, highlighting the region's distinct approach to AI governance.

Education challenges 

UK universities are working to balance AI literacy with academic integrity. Rather than banning AI tools outright, many institutions are developing nuanced policies that prepare students for an AI-enabled workforce while maintaining academic standards.

Different approaches

This approach differs from the U.S. model of rapid, market-driven deployment. While American tech giants push ahead with minimal oversight, Europe is attempting to build guardrails first. The EU AI Act introduces graduated requirements based on risk levels - an approach that could influence global standards much as GDPR shaped data protection.

Three metrics will indicate whether this measured strategy succeeds: Europe's ability to commercialize research, scale successful pilots, and build world-class AI companies. Early signs suggest the region may be able to compete but seems to be optimizing for careful oversight and practical applications rather than pure speed.

We publish daily research, playbooks, and deep industry data breakdowns. Learn More Here

One of the holy grails of unlocking the potential of AI is to build systems that can reason like humans. By improving AI's, Large Language Models in particular, ability to break down complex problems and apply logical steps.

Bagel’s research team has been exploring this problem. Analyzing LLM building techniques, especially fine-tuning techniques, to allow Large Language Models to evolve from pattern-recognizing prediction agents to true cognitive agents

Read More Here 🥯 

AI generated art

A look at the art and projects being created with AI.