AI

May 10, 2025

Published 1 month ago

TL;DR

DOJ/FTC seek Google Chrome divestiture; xAI launches Grok 3.5; SoundCloud faces AI copyright backlash.


Highlights

  • DOJ and FTC propose remedies against Google ’s search monopoly, including Chrome divestiture, mandated data sharing, and ending the $20B Apple search default deal; Anthropic warns this could deter AI investment 1.
  • Texas AG secures a record $1.375B privacy settlement from Google over alleged user data violations, increasing pressure on tech firms’ data practices 5.
  • xAI prepares Grok 3.5 beta with a 1M token context window, first-principles reasoning, and custom instructions; model selection removed 2.
  • Tsinghua’s “Absolute Zero” LLMs outperform prior models by 1.8 points via reinforced self-play reasoning, eliminating need for human-labeled data 3.
  • SoundCloud updates terms to allow AI training on user-uploaded audio, drawing criticism from 400+ artists and renewed calls for copyright safeguards and opt-out rights 4.
  • Over 400 UK artists, including Elton John and Paul McCartney, urge government action to require disclosure and licensing for AI training on copyrighted works 8.
  • AI tool using selfies predicts cancer survival with 80% accuracy, outperforming oncologists and highlighting advances in AI healthcare diagnostics 7.
  • Rippling raises $450M at a $16.8B valuation amid legal dispute with Deel, underscoring investor interest in AI-driven HR SaaS 10.
  • Omada Health files for $100M IPO with 38% revenue growth, reflecting renewed momentum in digital health and AI-enabled care 14.
  • Elon Musk’s Grok criticized for generating scantily clad AI images on user request, raising concerns over AI content moderation standards 11.

Commentary

Regulatory and legal scrutiny of AI and big tech remains front and center. The DOJ and FTC’s proposed remedies against Google —including a possible Chrome divestiture and mandated data sharing—signal a willingness to directly intervene in platform and data advantages that underpin both search and AI development 1. Anthropic’s warning that these measures could chill AI investment, especially for smaller companies, highlights the risk of unintended consequences for the broader AI ecosystem 1. Separately, Google ’s $1.375B settlement with Texas over privacy violations further underscores the rising compliance bar for data collection and use, especially for firms leveraging user data in AI models 5.

AI research and product development continue to accelerate. xAI’s Grok 3.5, with a 1M token context window and first-principles reasoning, reflects ongoing competition among advanced LLMs 2, while Tsinghua’s “Absolute Zero” paradigm demonstrates the potential of self-improving models that don’t require human-labeled data 3. These advances could reduce data bottlenecks and lower costs for model scaling, but also introduce new oversight and safety challenges as models become more autonomous 3.

Content rights and training data provenance are under increased scrutiny. SoundCloud ’s updated terms enabling AI training on user-uploaded audio 4, combined with high-profile artist advocacy in the UK for disclosure and licensing 8, point to likely regulatory tightening around copyrighted data in AI training. AI practitioners should expect more explicit consent, transparency, and compensation requirements in future data pipelines 4.

The enterprise and healthcare sectors remain active. Rippling’s $450M raise and Omada Health’s IPO filing signal sustained investor appetite for AI-driven SaaS and digital health, despite ongoing legal and competitive pressures 1014. In healthcare, AI models continue to show tangible value, with new tools outperforming clinicians in cancer survival prediction 7. Meanwhile, content moderation challenges persist, as seen in criticism of Grok’s image generation capabilities, emphasizing the need for robust guardrails in AI deployments 11.

Subscribe to AI Brief

Get daily ai updates delivered to your inbox

AI | Archive | Market Brief | Market Brief