AI-Powered news roundup: Edition 11
We’ve surpassed 5,000 subscribers! We appreciate your ongoing support!
Our bi-weekly AI news roundup is designed to keep you informed on the latest (and most important) developments, all in under 5 minutes.
In this edition:
1. Man faces charges for $10M streaming scam with AI-generated songs
Source: Forbes
In an unprecedented case, Michael Smith, a North Carolina musician, has been charged with orchestrating a $10 million music streaming fraud using AI-generated songs. According to federal authorities, Smith created hundreds of thousands of songs with the help of AI and streamed them using bot accounts on platforms like Spotify, Amazon Music, Apple Music, and YouTube Music. By inflating the streams, he generated millions in illegitimate royalty payments.
The indictment, filed in U.S. District Court in New York, details how Smith, along with unnamed co-conspirators—including an AI music company CEO—used fake artist names and song titles to avoid detection. Smith’s email communications reveal a carefully crafted strategy to bypass anti-fraud measures on streaming platforms.
Law enforcement called the scheme a "brazen fraud" and a first of its kind in the music industry. Smith now faces charges of wire fraud conspiracy, wire fraud, and money laundering conspiracy, each carrying a maximum 20-year prison sentence. Prosecutors emphasized that the royalties stolen were rightfully due to musicians, songwriters, and rights holders whose work was lawfully streamed.
2. OpenAI launches o1 series with advanced reasoning abilities
Source: ZDNet
OpenAI has introduced its new o1 series, designed to excel in complex problem-solving areas like science, coding, and math. Unlike previous models, the o1 series takes more time to think before responding, improving its ability to handle intricate tasks. OpenAI claims this refined reasoning process allows o1 to better strategize and identify mistakes, making it highly effective in fields requiring advanced reasoning.
A standout achievement of o1 is scoring 83% on the International Mathematics Olympiad (IMO) qualifying exam, a significant improvement over GPT-4o’s 13%. Despite its strengths in reasoning, the o1 series lacks some features of GPT-4o, such as web browsing and media uploads, making GPT-4o more suited for general use.
In addition to the full-sized o1 model, OpenAI has also launched the o1-mini version, which is 80% cheaper and optimized for coding tasks, offering a faster and more cost-effective option for developers. While o1 is still an early release and has limitations, it showcases OpenAI's focus on advancing the depth of AI's thinking capabilities.
3. Microsoft launches Copilot Pages for real-time AI collaboration
Source: The Verge
Microsoft has introduced Copilot Pages, a new feature within its 365 Copilot suite designed for collaborative AI-powered work. Copilot Pages acts as a canvas for "multiplayer AI collaboration," enabling users to interact with the Copilot chatbot and collaboratively edit responses in real-time. This innovative work model allows users to pull data from files, the web, and other sources, facilitating tasks such as creating project plans, meeting notes, and business pitches.
Jared Spataro, Microsoft’s VP of AI, describes Copilot Pages as a revolutionary way for teams to interact with AI and each other, marking a new era of human-to-AI collaboration. Users can easily share and edit pages via links, similar to shared Word documents, and embed them within other pages. The feature is integrated with Microsoft’s BizChat, allowing users to pull data seamlessly from their work files.
Copilot Pages is being rolled out to Microsoft 365 Copilot customers and will soon be available to all subscribers. Additionally, Microsoft has launched Copilot agents, AI-powered virtual colleagues designed to automate tasks, such as managing emails or data entry. With Copilot Studio, users can now create custom agents to assist in platforms like SharePoint and Teams, enhancing productivity across a variety of workflows.
4. Slack bets on AI features
Source: TechCrunch
Slack is evolving into a "work operating system" with AI integration, according to CEO Denise Dresser. The platform, now owned by Salesforce, has introduced new AI features, including AI-generated Huddle summaries and seamless integration with Salesforce’s AI agents. Users will also be able to leverage third-party AI tools for web search and image generation directly within Slack. These features are available to those on a pricier Slack AI tier.
Dresser envisions Slack as more than a messaging platform, positioning it as a digital workspace where people can collaborate with AI in the flow of work. New tools like Agentforce allow Salesforce customers to perform real-time data analysis within Slack, while Cohere, Anthropic, and Adobe offer AI agents that enhance productivity.
While some CEOs, like Klarna’s, are moving toward in-house AI solutions, Dresser emphasizes that trust and safety are paramount in Slack’s AI integrations. Addressing privacy concerns raised earlier this year, Slack reassured users that no customer data is used to train its large language models (LLMs). As Slack continues to integrate AI, it faces the challenge of balancing innovation with user skepticism over privacy.
5. Anthropic unveils Claude Enterprise plan for AI-powered collaboration
Source: Anthropic
Anthropic has launched the Claude Enterprise plan, designed to help organizations securely integrate AI into their workflows. With a 500K context window, the plan allows teams to work with large datasets, such as entire codebases, or hundreds of pages of documents.
Claude's new GitHub integration enables engineering teams to collaborate directly with code, making it easier to debug, iterate, and develop features:
In addition to its collaboration tools, the Enterprise plan prioritizes data security with features like Single Sign-On (SSO), role-based permissions, and audit logs. Anthropic also assures that Claude does not train on customer conversations or content, providing peace of mind for companies concerned about privacy.
The expanded context window and GitHub integration make Claude a powerful tool for scaling expertise across projects and teams. Early adopters like GitLab and Midjourney are using Claude for tasks ranging from brainstorming and content creation to coding and policy iteration. By combining enterprise-grade security with deep AI collaboration, Claude helps teams produce higher-quality work faster and more securely.
To get a quote on the Enterprise plan, contact Anthropic sales here: Claude Enterprise.
5. Greenlite secures $28.5M to revolutionize construction permitting
Source: TechCrunch
Greenlite, a New York-based startup, has raised $28.5 million in a Series A round to streamline the often cumbersome construction permitting process. Co-founded by Ben Allen and James Gallagher, Greenlite aims to standardize permitting across thousands of jurisdictions, offering a software solution that accelerates plan reviews and inspections for developers, retailers, and quick-service restaurants. The platform integrates construction drawings, zoning data, and building codes, making it easier for customers to navigate complex regulatory landscapes.
Greenlite stands out by providing faster, more transparent permitting, and is authorized to conduct reviews and inspections in approximately 2,000 jurisdictions. Unlike competitors such as PermitFlow and Accela, which focus on application submissions, Greenlite emphasizes full-service digital plan reviews. This approach has attracted clients from diverse sectors, including hospitality, industrial, and green infrastructure.
The new funding, led by Craft Ventures, will be used to fuel Greenlite’s expansion and enhance customer acquisition. With a team of 30 employees, the company plans to grow further by year’s end, capitalizing on the perpetual demand for building permits across various industries.