AI Daily Briefing — 2026-06-10

OpenAI has taken a major step toward going public, confidentially filing for an IPO on Monday. The move positions the artificial intelligence leader for what could be one of the most significant stock market debuts in recent years. However, the same week brought a contrasting development for CEO Sam Altman's other enterprise, Tools for Humanity, which is reportedly conducting layoffs according to a Business Insider report. TechCrunch AI has reached out for confirmation.

Tools for Humanity is best known for World, a verification project that uses a distinctive silver orb to scan individuals' irises. The company aims to distinguish human activity from bots in an increasingly automated digital landscape. It also plans to leverage these biometric scans to facilitate transactions with its own cryptocurrency, Worldcoin. These ambitions previously helped the startup secure a $2.5 billion valuation from investors including Andreessen Horowitz and Bain Capital.

Despite partnerships with U.S. firms like Tinder and Zoom, the company now faces downsizing amid revenue challenges. Internationally, Tools for Humanity has encountered regulatory pushback. In Kenya, operations were banned over privacy and financial concerns, while South Korea imposed an $830,000 fine for alleged privacy law violations. The company had offered individuals in several countries roughly $50 in Worldcoin in exchange for their biometric data, a practice that drew widespread scrutiny.

As the use of AI agents is expected to surge by up to 300% in the next two years, corporate leaders are grappling with the shift toward a hybrid human-AI workforce. Unlike traditional automation, these agents can autonomously manage complex tasks across multiple systems, acting more as collaborators than tools. Early adoption in customer service, HR, and sales has already boosted productivity by 30-50%, prompting a fundamental rethink of workplace roles and culture. Over three-quarters of HR leaders believe AI agents will transform workplace norms, with 86% of chief HR officers predicting that managing this digital labor will become central to their role.

This transformation demands a mindset shift, says Ateet Jayaswal of Wipro. He emphasizes that leaders must focus on reskilling employees for higher-value work as AI takes over routine tasks. For example, Wipro deployed a custom AI agent to handle 50 HR tasks, slashing query response times from 48 hours to five seconds. This freed human employees to focus on creative, collaborative problem-solving. However, Jayaswal warns that humans must remain in the loop, especially when AI handles sensitive data, requiring robust governance and guardrails.

With an estimated 75% of roles needing redesign or reskilling by 2030, leadership must prioritize change management. Jayaswal stresses that navigating this shift fluently will be key to unlocking AI's full potential. As agentic AI evolves, establishing data privacy rules and governance layers, like an AI council, will be critical to ensuring safe and effective integration into enterprise systems.

Anthropic has launched Claude Fable 5, the first publicly accessible version of its advanced Mythos AI model. The model, previously limited to a small group of partners due to cybersecurity concerns, is now available through Anthropic's API and enterprise plans. Fable 5 excels in software engineering, knowledge work, and vision tasks, but comes with strict safety guardrails. In high-risk areas like cybersecurity and biology, it blocks responses and defaults to the less powerful Claude Opus 4.8.

The rollout includes a phased subscription approach. Through June 22, Fable 5 is included in Pro, Max, Team, and Enterprise plans at no extra cost. After that date, it will require usage credits until it is restored as a standard feature. Anthropic has also deployed Mythos 5 to pre-approved organizations. The launch follows the company's warning that AI systems may soon achieve recursive self-improvement, and its call for a coordinated safety brake across major AI labs.

To prevent misuse, Anthropic stress-tested Fable 5 with over 1,000 hours of internal and external red-teaming, finding no universal jailbreaks. However, the company acknowledges novel attacks remain possible. As a safety measure, Anthropic now requires a 30-day data retention on all traffic, even for enterprises with previous zero-retention agreements. The data will not be used for training, only to defend against attacks and reduce false positives.

Early testing shows strong performance. Analytics firm Hex reported Fable 5 scored 90% on complex analytical tasks, while other partners praised its ability to build full applications and handle UI design. Pricing is set at $10 per million input tokens and $50 per million output tokens.

The era of FAANG is giving way to a fresh acronym: MANGOS. As SpaceX gears up for a record-breaking IPO on Friday, with Anthropic and OpenAI also preparing their own public offerings, the tech landscape is shifting. The new grouping—Meta, Anthropic, Nvidia, Google, OpenAI, and SpaceX—reflects the rising dominance of AI and space ventures over traditional streaming and e-commerce giants.

European startup Lovable has announced it now generates $500 million in annualized revenue, a significant jump from the $400 million figure it reported just a few months ago in February. The company, which was founded in late 2023 and has yet to celebrate its third anniversary, claims users are creating over one million new projects each week, bringing the total number of projects built on its platform to more than 50 million.

According to Lovable, its user base is largely non-technical, consisting of founders, designers, and salespeople who are building everything from websites and e-commerce storefronts to internal tools like CRMs and HR platforms. The company suggests this trend signals a shift away from traditional SaaS software, as users opt to "vibe code" their own solutions instead of purchasing expensive annual contracts.

However, questions remain about the long-term viability of vibe-coded software. While building such applications is relatively easy, maintaining them over time poses a significant challenge. Software relies on a constantly changing ecosystem of dependencies and third-party services, meaning even well-written code can break frequently. The true test for Lovable and similar platforms will be whether users stick with their projects or abandon them as maintenance demands grow.