NerdNews

The US government has imposed export controls on Anthropic's AI model Fable 5, citing security concerns. The model was taken offline after the government gave Anthropic a 90-minute ultimatum to restrict access to the model. The situation highlights the need for a more considered regulatory approach to AI development and deployment.

Adobe has introduced new capabilities for its Firefly AI assistant, including a 'reimagined' AI studio that allows for editing and generating designs from a single interface. The Firefly AI assistant can now generate brand logos, assemble video clips into draft cuts, and create storyboards. The updates aim to improve design consistency and make projects easier to organize.

Microsoft Scout is a new enterprise autopilot built on the open-source agent framework OpenClaw, integrating with Work IQ and providing autonomous task execution with a focus on security and governance.

Anthropic's dynamic workflows feature makes Claude more reliable by allowing it to orchestrate large, multi-agent tasks and assign tasks to many subagents, increasing the quality of results. This feature has made previous workarounds obsolete, highlighting the rapid pace of AI development.

Adobe has introduced new AI features, including Elements, which allows for consistent character and object generation, and Projects, which keeps all generations and assets together. The company has also launched an AI Assistant in Photoshop, Premiere, and other tools, designed to respond to specific creative tasks.

GitGuardian improved dashboard performance with 5 PostgreSQL optimizations: late joining, asynchronous counting, premium replication, full-text search optimizations, and denormalization, resulting in a p95 of 1s.

Netflix has detailed a system for scaling camera file processing, used across global film and TV production workflows. The system processes large volumes of raw camera footage, enabling ingestion, validation, metadata extraction, and transformation into standardized formats.

Elastic has agreed to buy DeductiveAI, an AI startup that catches and resolves bugs in software, for up to $85 million. The acquisition reflects a trend of tech incumbents buying AI-native startups to integrate their technologies into existing product suites.

Baseten, an AI inference company, is reportedly raising $1.5 billion at a $13 billion valuation, just months after its last funding round. The startup promises to handle inference quickly while controlling costs by routing requests to the best-for-task model.

TechCrunch highlights 11 startups from YC's Demo Day, featuring companies like 9 Mothers, Arga Labs, and Adialante, which are working on AI-powered counter-drone systems, digital twin environments, and mobile MRI clinics. These startups have garnered attention from investors due to their innovative solutions and potential for growth.

The US government has suspended Anthropic's AI models, Claude Fable 5 and Mythos 5, due to national security concerns. This move has raised concerns about the UK's reliance on US tech and calls for AI sovereignty. Anthropic has disputed the claims, stating the vulnerability is minor and has pushed back against the US government's decision.

Athena, an industry coalition, aims to use AI to find and fix vulnerabilities in open-source software before attackers can exploit them. The coalition, founded by Chainguard, includes over two dozen members such as financial institutions and infrastructure vendors. Athena focuses on libraries, containers, and other components that underpin web browsers, data centers, smartphones, and payment systems, and has already made significant progress in its first month of operation.

The White House has ordered Anthropic to shut down its advanced AI models, Claude Mythos and Fable 5, due to concerns over export controls and national security. The move has sparked debate over AI regulation, with some arguing that the government is making up rules as it goes along. Anthropic claims it did not violate any concrete procedures or rules, while the White House contends that the company behaved recklessly.

Midjourney, known for its AI image generator, is developing a full-body ultrasonic scanner that can scan a body in 60 seconds. The scanner uses half a million squares to emit ultrasonic waves and record ripples, creating a 3D map of the body. Midjourney is building spas to house the scanners and plans to expand globally, aiming to have 50,000 scanners available by 2031.

The Federal Energy Regulatory Commission (FERC) has ordered grid operators to fast-track interconnection requests from data centers and other large electricity users, aiming to address the growing demand for electricity from AI data centers. This move is expected to help tech companies and developers connect to the grid in a timely manner, reducing their reliance on expensive and complicated on-site power solutions.

Uber plans to launch a premium robotaxi service in Houston by mid-2027, in partnership with Lucid and Nuro. The service will feature Lucid Gravity SUVs equipped with Nuro's self-driving system and will compete with Waymo's commercial robotaxi services in the city.

Amazon employees who testified at Seattle City Council hearings about data center regulations are facing disciplinary action, including possible termination. The employees, who are members of Amazon Employees for Climate Justice, allege that Amazon is retaliating against them for exercising their right to free speech. The incident highlights the controversy surrounding data centers and their impact on local communities.

Waymo has issued a software recall for over 3,800 robotaxis due to a risk of driving into freeway construction zones. The recall was prompted by several incidents in Phoenix and the San Francisco Bay Area. Waymo is working on a software fix to help the cars detect and avoid construction zones.

Waymo is recalling over 3,800 self-driving taxis due to a software issue that could cause them to enter closed freeway construction zones at speed. The company is working on a fix and has restricted freeway driving. This is the second recall for the Alphabet-owned robotaxi company in a bit more than a month.

Waymo has recalled nearly 4,000 robotaxis to restrict them from driving on highways due to issues with navigating construction zones. The company has identified 13 instances of its robotaxis driving into closed highway sections and is working on a fix.