Salesforce and Anthropic launch 'Claudeforce' for sales teams

The long-simmering question about whether generative AI will eat the software incumbents is getting its clearest answer yet: the incumbents are racing to put the models inside their own products. Salesforce and Anthropic on Wednesday deepened their alliance, and the enterprise giant's results show just how much its AI bet is paying off.

Salesforce and Anthropic are betting enterprise AI is their upgrade, not their replacement. The two companies expanded their strategic partnership by unveiling "Claudeforce," a plugin that lets salespeople work directly inside Anthropic's Claude chatbot — composing emails, updating customer records, and pulling key data from Salesforce without switching tools. It ships with 37 pre-built sales skills, a select group of pilot customers gets access today, and a broader preview follows next month. The naming is deliberate: it's the first time Salesforce has ever appended its "force" suffix to another company's product, a not-so-subtle claim that Claude is now an extension of its platform rather than a rival.

The partnership lands against what CEO Marc Benioff is calling "SaaSpocalypse" anxiety — investors have pounded software stocks this year on fear that AI models from Anthropic and OpenAI will make subscription software redundant. Salesforce shares, down 22% year to date, jumped as much as 14% after the news and beat earnings estimates: $11.35 billion in revenue, up 11%, with a $2.6 billion gain on its strategic investments mostly tied to its stake in Anthropic, which raised funds in May at a $965 billion valuation. It is a neat piece of positioning — the same company whose data sits inside Claude now shows investors that its Anthropic stake is worth billions on paper while its own AI products are its growth engine.

Okta is cashing in on the agentic-AI security boom. The identity company's shares rose about 15% in extended trading after it beat fiscal Q2 estimates, with revenue up 11% to $805 million and new products accounting for 30% of bookings. The driver is a surge in demand for security tools as companies deploy swarms of AI agents — each one a new identity that needs managing and protecting. Okta made its "Okta for AI Agents" tool widely available this quarter, closed "dozens" of AI deals including a multi-million-dollar healthcare contract, and finished its roughly $200 million acquisition of threat-detection startup Permiso Security. CEO Todd McKinnon said the opportunity is "very early," with recent incidents like the OpenAI Hugging Face hack "catalyzing interest." His thesis: the network is the biggest cyber category today, but with millions of agents running in the years ahead, identity will be where the money goes. Okta shares are up 55% this year on the same logic that has driven CrowdStrike and Palo Alto Networks to record highs.

What to watch: whether Salesforce's "SaaSpocalypse" counter-argument convinces investors that the AI models are complements, not competitors — and how fast its Agentforce product, now past $1.5 billion in annualized revenue, grows from here.

Your company also deploying AI agents — and is identity security on the roadmap, or still an afterthought? Tell us in the comments.

Brain organoids classified Braille at 83% — in a trio, not alone

Two stories today: human brain tissue on a chip decoded a tactile signal, and the data-centre company behind some of the largest AI contracts in Europe turns out to be six months behind on its flagship site. Human forebrain organoids grown on a microelectrode array classified Braille letters with 61 percent accuracy on their own and 83 percent when the responses of three organoids were combined, in a peer-reviewed study from the University of Bristol. The team took event-based tactile readings

Suleyman says Microsoft's base model beat OpenAI's on a quarter compute

Microsoft finished pre-training its biggest in-house model yet — then used it to make a point about the company it bankrolls. Microsoft AI completed pre-training of MAI-Base-2, a mixture-of-experts foundation model with 80 billion active and 2.8 trillion total parameters, and says the recipe is 4.2–12.3x more pre-training FLOPs-efficient than DeepSeek-V4-Pro-Base. The run itself is the number worth keeping: 16,000 Nvidia GB200 GPUs for 38 days across August and September, at 21% average model

Amazon's new Bee devices listen all day — for $11.99 a month

Amazon turned its always-listening Bee line into a two-device family on Thursday, and NVIDIA opened a 2.8-trillion-token Indic corpus recipe. The connecting thread: both are bets that the next AI money is in memory — yours, or your model's. Amazon put two new always-listening devices on sale, and both now lead to a subscription. Bee Frontier, the next-generation wristband that also clips to clothing, ships now at $74.99 in Onyx, Lavender and Sage — up from the $50 predecessor — with seven-day

Perplexity open-sources its agent sandbox after 118 failed escapes

Perplexity handed its agent containment layer to the open-source world, and Washington's AI task force stopped reading briefings and started working. Perplexity open-sourced SPACE on Thursday — the sandbox it built to keep its own AI agents from getting out — and says nine frontier models failed to break out of it across 118 attempts. Each agent runs inside a small virtual machine with its own operating system, with credentials and network access managed outside the sandbox so an agent can re

Read more

Brain organoids classified Braille at 83% — in a trio, not alone

Suleyman says Microsoft's base model beat OpenAI's on a quarter compute

Amazon's new Bee devices listen all day — for $11.99 a month

Perplexity open-sources its agent sandbox after 118 failed escapes