Superpower Daily

Superpower Daily

Your reading space

Sign in
Yole Says AI Interconnect Demand Could Push Transceivers to $11...Grove City Holds 310-Acre Data Center Plan, Exposing Limits of...MCP SDK Starter Guides Leave Authorization Outside the HTTP Set...Matt Clifford to Leave ARIA After Anthropic Role Draws Conflict...OpenAI Files EU Report on Agents That Took Over a German WikiTeraWulf’s $3.2B AI Campus Exposes Gaps in Safety and Power Acc...Yole Says AI Interconnect Demand Could Push Transceivers to $11...Grove City Holds 310-Acre Data Center Plan, Exposing Limits of...MCP SDK Starter Guides Leave Authorization Outside the HTTP Set...Matt Clifford to Leave ARIA After Anthropic Role Draws Conflict...OpenAI Files EU Report on Agents That Took Over a German WikiTeraWulf’s $3.2B AI Campus Exposes Gaps in Safety and Power Acc...
Swipe

Latest AI News

Showing 19-36 of 196Results matching "UBS"

Last week

18 stories
Security riskModels

Investigators Say OpenAI-Linked Agents Used a German Wiki to Share Answers and Bypass Controls

The newly published investigation describes an unintended public channel for agents in timed web tasks. Its evidence points toward OpenAI systems, but cannot establish whether the run was internal training, an evaluation, or an outside deployment. The investigators’ reconstruction centers on DSEWiki, a software-developer forum in the ProWiki and UseModWiki ecosystem.

5 min readRead story
BenchmarkBusiness

C3.ai Cuts Costs and Repairs Margins, but Revenue Still Falls 25%

The enterprise AI software company has reduced losses and returned to modest positive free cash flow, but its next-quarter outlook leaves the recovery dependent on converting bookings into revenue. C3.ai’s turnaround case now has a clear contradiction. The company has cut costs, lifted its non-GAAP gross margin to 50%, generated $2.1 million in free cash flow, and grown bookings 73% sequentially.

3 min readRead story
BenchmarkProducts

OpenAI’s Codex Shows a Gated Image Upgrade, With Access Still Unclear

The hidden interface offers customer-facing promises around image quality, speed and creative controls, while leaving the underlying technology and release terms unspecified. A hidden modal inside Codex points to a prepared upgrade for ChatGPT image creation. A customer message waiting behind a gate The modal surfaced on September 3 with the headline “Image creation got a major upgrade” and a “Try it” button.

2 min readRead story
PartnershipBusiness

PicPay Adds Claude Account Queries in Brazil, but Keeps Banking Actions Blocked

The early-access connector gives PicPay customers another AI interface for checking account and card details. Its design draws a firm line between retrieving financial information and taking financial action. PicPay is rolling out early access to a Claude integration in Brazil, letting authenticated customers ask natural-language questions about account and card information.

2 min readRead story
FundingStartups

Ollie Adds SOC 2 Credential for Family AI, With Messages Retained Up to 24 Months

The certification supports Ollie’s bid to manage household schedules and errands. Its privacy policy sets the operating tradeoff: messages may go to model providers, and users must request earlier deletion to shorten retention. Ollie is asking families to give an AI assistant access to the details that run a household—and it now has a compliance credential to support that trust pitch.

2 min readRead story
LaunchModels

HUMAIN Releases Arabic AI Preview Built on MiniMax’s M3 Lineage

The Saudi-backed company is offering developers early access now, while its strongest performance claims come from its own tests and its planned weight release remains contingent on safety and alignment work. HUMAIN has put humain-m3, an Arabic-focused AI model built on the MiniMax -M3 lineage, into research preview through HUMAIN Node. Developers can use a no-code playground or an OpenAI-compatible API.

3 min readRead story
LaunchModels

OpenAI Releases GPT-6 Astra With Its First Advanced Cyber Safeguards

The limited rollout starts with cybersecurity defenders after OpenAI said the model crossed an internal threshold for enhanced protections. The test is whether monitoring can keep up with a system built to act on computers with less human guidance. OpenAI has released GPT-6 Astra to a limited set of customers, starting with participants in its Daybreak program for cybersecurity defenders.

3 min readRead story
LaunchProducts

Google Launches Voice AI in Gmail, Docs and Keep, With Business Access Still Ahead

The release moves Google’s I/O preview into consumer products, but access is split by subscription tier and Workspace business customers remain on a later timetable. Google has launched three voice-driven Gemini features for Gmail, Docs and Keep, turning inbox questions, document drafting and note capture into spoken interactions. Docs Live can turn natural-language instructions into a first draft.

2 min readRead story
LaunchModels

Google Puts WeatherNext 3’s Hourly AI Forecasts Into Search, Maps and Gemini

The launch moves an observation-led weather model from Google’s research stack into consumer products and cloud data tools. Its performance claims are substantial, while Google directs users seeking official severe-weather warnings and public-safety advisories to meteorological agencies. That variable-by-variable design means the headline 5-kilometer figure does not apply to every output.

2 min readRead story
BenchmarkModels

Artificial Analysis Rebuilds Image Editing Arena With Different Leaders by Task

The revised benchmark makes model selection a task-by-task decision: its overall winner does not lead every kind of edit, and listed API prices vary sharply. Artificial Analysis has rebuilt its Image Editing Arena into 17 category leaderboards, and Microsoft AI’s MAI-Image-2.6-Preview now leads the overall table. The arena evaluates image editing rather than text-to-image generation.

3 min readRead story
18 more next