TRO | Salesforce And Anthropic Launch 'Claudeforce'

Plus, Nvidia Agrees To Acquire Hugging Face For US$12.9 Billion and A London AI Lab Just Outperformed Claude and GPT

Author

Meliora Company
August 31, 2026 • Estimated Reading Time: 4 minutes

 Subscribe | 31st August, 2026

In this fast-moving Vertical AI economy, headlines are everywhere but optimism is rare.

Here’s our take on 3 stories that will help you be relentlessly optimistic about the future.

1. Salesforce and Anthropic Launch ‘Claudeforce’ As The ‘SaaSpocalypse’ Fails To Materialise

Anthropic and Salesforce have launched Claudeforce, an expanded strategic partnership that embeds Claude directly into the Salesforce platform and makes Salesforce natively accessible from within Claude. The launch product is Salesforce in Claude, a plugin with 37 prebuilt sales skills covering meeting prep, deal health review, and pipeline management. Claude is simultaneously the default model powering Agentforce, Slackbot, and Claude Code across Salesforce's entire workforce. Nine of the top ten AI companies now use Salesforce and Slack, with their combined platform spending up 435% year on year. Salesforce shares surged 13% on the news.

Here’s why this matters: For most of 2026, a significant wave of market pessimism built around a single thesis: that AI agents would hollow out the software businesses underneath them. Benioff's response, delivered with a 13% stock surge and a record earnings beat, was direct: frontier models depend on CRM, they do not replace it. Claudeforce makes that argument in product form. Claude needs Salesforce's decades of customer data to act on. Salesforce needs Claude's reasoning to activate it. The most valuable enterprise AI deployments are not replacing existing systems. They are being built on top of them.

2. Nvidia Agrees To Acquire Hugging Face For US$12.9 Billion

Nvidia has agreed to acquire Hugging Face, the open-source AI model hub used by millions of developers worldwide, for US$12.9 billion, a deal that would mark a near tripling of Hugging Face's last known valuation of US$4.5 billion in 2023. The acquisition gives Nvidia a dominant foothold in the open-source AI ecosystem at a moment when the biggest closed-source labs, including OpenAI, Google, Amazon, and Anthropic, are building their own chips to reduce reliance on Nvidia hardware. Hugging Face generates approximately US$150 million in annual revenue and was approaching profitability.

Here’s why this matters: Nvidia's move on Hugging Face is a defensive play dressed up as an offensive one. Its chip dominance is being threatened from two directions simultaneously: closed-source labs building proprietary silicon to escape Nvidia dependency, and Chinese open-source models offering competitive capability at a fraction of the cost. Acquiring Hugging Face addresses both. A thriving open-source ecosystem keeps more of the market dependent on Nvidia hardware rather than proprietary chips, while giving Nvidia a distribution platform for its own open-source models. The acquisition also hands Nvidia a re-entry into cloud computing, with Hugging Face's existing developer infrastructure providing a ready-made customer base for idle compute capacity.

3. A London AI Lab Just Outperformed Claude And GPT While Using A Fraction Of The Compute

Inherent, a London AI lab founded by four Google DeepMind alumni, has released Faraday, an AI agent that outperformed Claude Opus 4.8 and GPT-5.5 at independently replicating published scientific research papers, using a model 27 billion parameters in size against frontier models many times larger. Faraday was evaluated not just on accuracy but on what Inherent calls "research taste": the instinct for which experiments are worth running and how to design them well. The company emerged from stealth in May with a US$50 million seed round and operates out of King's Cross, London, with a team of 12 planning to grow to 25 by year end.

Here’s why this matters: The capability benchmark here is less interesting than the efficiency one. Outperforming frontier-scale models on a specialist research task using a model a fraction of their size suggests the AI research race is no longer solely a question of who has the most compute. Using reinforcement learning to develop generalised scientific reasoning rather than training narrowly on how science is conducted, is a meaningful architectural bet. The broader implication is that domain-specific AI agents built for precision tasks are beginning to outperform general-purpose models in ways that matter commercially. The era of one frontier model for everything is being challenged from multiple directions simultaneously, and London is increasingly where that challenge is being mounted.

From the Spotify Mafia to the Lovable Mafia: Inside Stockholm’s Founder Network

Vibe-coding darling Lovable just raised US$400 million at a US$13.3 billion valuation. But it isn’t the only Stockholm startup putting up huge numbers lately.

Unpack what’s driving startup growth in the Nordic ecosystem, from Sweden’s social safety net to the wave of US capital flowing in.

Still Curious?

Our Pursuit Of Better

Stability AI has raised US$76 million in Series B funding from an unusual roster of backers: Universal Music Group, Sony Music Group, Warner Music Group, Electronic Arts, AMD Ventures, and Pacific Alliance Ventures, bringing total funding to US$232 million. The round positions Stability less as a consumer AI company and more as a purpose-built creative production platform for music, gaming, and entertainment professionals.

Perceptron, founded by two former Meta AI research scientists, has raised US$21 million to build general-purpose vision models for industrial robots. Its latest model, Isaac 0.5, gives warehouse and factory robots the ability to perceive, reason, and act flexibly across complex physical environments without being limited to a single repetitive task. The model is open-weight and targets manufacturing, logistics, security, and media and entertainment.

Atorie, the AI-powered fashion start-up connecting luxury manufacturers directly with consumers, has raised US$9.5 million in seed funding. The platform sells goods made from the same materials and factories as high-end brands at a fraction of the price, using AI agents to predict demand, manage factory supply, and personalise the shopping experience. The company is tracking toward US$55 million in annualised revenue this year.

The Pursuit of Better, Together

If this week’s newsletter sparked something, pass it on to a colleague. The best conversations often start with “You’ve got to read this”.