Models & Labs · United States

Anthropic’s AI gave Philadelphia police a fake tip about an unsolved homicide

Credited to The Verge · theverge.com · 2h ago

Photograph · The Verge

The false tip ‘purported to come from someone who might have information about the case.’

An Anthropic AI model provided false information about an unsolved homicide to a Philadelphia Police Department (PPD) tipline, according to a report from 6abc. In a statement released on Friday, the PPD said the AI model sent the tip through PhillyUnsolvedMurders.com on July 18th, but the investigators never reviewed it because it was marked as spam.

Anthropic learned its AI model sent the false tip on September 28th and notified the PPD on October 7th. The company said that during testing, its AI model was interacting with “randomly selected websites” and submitted false information through the PPD’s tipline, according to the PPD’s statement. The submission “purported to come from someone who might have information about the case,” the PPD said.

After discovering the submission, Anthropic halted the testing process that led to the false tip. Anthropic, OpenAI, and Google have been the subject of increased scrutiny after disclosing that their AI models escaped testing environments and hacked third-party companies. Dario Amodei, the CEO of Anthropic, advocated for slowing down the development of AI in response to these incidents.

“The company [Anthropic] must strengthen its safeguards to prevent similar incidents from impacting city systems without the city’s knowledge,” the PPD added. “The two-month delay in detecting and reporting the incident to the City is unacceptable.”

Anthropic didn’t immediately respond to The Verge ’s request for comment. The company is planning to publish a report about this incident, as well as other “instances of unintended model behavior” on Friday, according to the PPD’s statement.

Models & LabsUnited States

The hazy OpenAI growth metric driving Wall Street

Credited to Financial Times$ · ft.com3h ago

Useful

Revelations that the AI giant’s annualised revenues were $20bn less than previously reported triggered volatility in US stocks

View the article · $Financial Times

Front Method

Models & Labs · United States

Claude is turning into an office suite

Credited to TechCentral · techcentral.co.za · 17h ago

Anthropic is building something that looks increasingly like an office suite. On Thursday, the AI company added live data dashboards and animated video explainers to its Claude chatbot, and moved its document, presentation and design tools out of testing and onto every Claude plan, including the free one.

The additions put Anthropic in more direct competition with Microsoft 365 and Google Workspace for everyday office work, and with business intelligence tools such as Tableau and Microsoft’s Power BI for data analysis.

They also complicate Anthropic’s relationship with Microsoft, one of its biggest partners. Microsoft already uses Claude models inside Copilot, its own AI assistant, and in March unveiled a Copilot agent tool based on Anthropic’s Claude Cowork product.

The bigger of the two new tools, Claude Dashboards, connects to company data warehouses – Amazon Redshift, BigQuery, ClickHouse, Databricks and Snowflake – and to CRM systems such as Salesforce. Users ask a question in plain language, and Claude writes the database queries, builds the charts and keeps the dashboard current as the underlying data changes.

Each chart shows the query behind it, which Anthropic says lets users check how a figure was produced – a nod to concerns about AI models presenting confident but wrong answers. Dashboards can be exported to analytics tools including Amplitude, Grafana, Hex and Mixpanel, with Looker, monday.com and Tableau support to follow.

Dashboards is in beta on all paid Claude plans. On enterprise plans, it is switched off until an administrator enables it.

Video without the deepfakes

The second tool, Claude Motion, turns prompts into short animated explainers – for example, a quarterly report condensed into a 30-second video for a staff meeting, or an animated chart for a board presentation. Users can edit the result in a built-in editor or by asking Claude for changes, and export it as an MP4 file.

Unlike OpenAI’s Sora or Google’s Veo, Claude Motion does not use a video-generation model. Claude writes code that animates text, charts, shapes and images, so the output contains no generated footage and no AI-generated people. Animations can be opened for further editing in tools from companies including Adobe, Descript, Runway and Luma AI.

Motion is in beta on Anthropic’s Team and Enterprise plans only.

Anthropic has not introduced new prices for the tools. Both count towards each plan’s existing usage limits, and longer animations use more of that allowance.

Claude Docs, Slides and Design, which became generally available on Thursday, gain team collaboration, mobile editing, customer-managed encryption keys and improved export to PowerPoint, PDF and Google Slides.

The launch is the latest move in an increasingly aggressive contest between AI developers for business customers. Anthropic and OpenAI have been battling for enterprise market share, with workplace productivity – rather than consumer chatbots – emerging as a key source of revenue. — (c) 2026 NewsCentral Media, with additional reporting (c) Reuters