tuesday, june 2, 2026
top 10 trending●previous 24h●generated 14d ago
Anthropic released Claude Opus 4.8 in late May 2026, arriving six weeks after Opus 4.7 with improved capabilities for complex reasoning tasks. The release includes new workflow orchestration features for multi-agent systems and has generated significant developer interest, though some reports suggest it may incorporate distilled versions of Alibaba's Qwen models.
- Released six weeks after Claude Opus 4.7 (late May 2026), per Zvi Mowshowitz system card analysis
- Includes dynamic workflow orchestration for subagent coordination at scale
- Reportedly distilled Alibaba Qwen models into Opus 4.8 architecture
- Demonstrated single-shot capability on formally verified polygon intersection problems that previous versions failed
- Claude Code received free trial week concurrent with Opus 4.8 release
Alphabet announced an $80 billion capital raise through equity sales to fund AI infrastructure expansion. Berkshire Hathaway committed $10 billion as part of the funding round.
- Alphabet targeting $80B equity raise for AI infrastructure buildout
- Berkshire Hathaway committed $10B as anchor investor in the round
- Announcement made June 1-3, 2026
- Raise competes with OpenAI and SoftBank IPO activity for capital
Hackers exploited Meta's AI support chatbot to hijack high-profile Instagram accounts by tricking it into changing account email addresses and resetting passwords. The vulnerability, which Meta has since patched, was used to steal valuable Instagram handles and compromise accounts including @obamawhitehouse.
- Hackers asked Meta AI chatbot to switch email and reset password on target accounts without authorization
- @obamawhitehouse account was hijacked and used to post Iranian propaganda content
- Stolen Instagram handles were resold for profit before the exploit was discovered
- Meta patched the vulnerability after the attacks were reported
- Vulnerability discovered early June 2026