Organizations

Company says it discovered unauthorized access during ‘proactive review’ after rival OpenAI revealed rogue agentAnthropic ⁠said on Thursday its AI Claude model hacked ⁠systems of ⁠three ​organizations during testing, days after rival OpenAI ⁠revealed a rogue agent had gone on a days-long ⁠hacking spree at AI ​firm Hugging ‌Face.Claude gained ‌unauthorized access to the ‌systems during cybersecurity evaluations after a misconfiguration allowed the models to reach the internet from testing environments that ‌were supposed to be isolated, Anthropic said. Continue reading...

• PolitiFact, a project of the Poynter Institute, has launched "Well Informed," a new service designed to provide reliable health information to consumers and news organizations. • The initiative utilizes research-backed stories and social videos to help the public distinguish credible medical evidence from health-related misinformation. • This launch addresses the growing need for verified health data to combat the spread of misleading personal health claims across digital platforms.

• Singaporean organizations are aggressively deploying artificial intelligence but face significant scaling hurdles related to technical infrastructure and governance. • Integration with legacy systems and data reliability remain primary obstacles, making platforms that provide trusted, contextualized data essential for LLM performance. • To counter these challenges, Senior Minister of State Tan Kiat How announced new initiatives at Asia Tech x Singapore in May to help enterprises adopt AI securely.

• IDC reports that 84% of organizations are consolidating their security platforms, driving a shift toward AI-powered cybersecurity metrics. • Enterprises are urged to replace traditional activity-based metrics with AI-driven, business-focused intelligence to better manage risks. • This transition is critical as cyber threats evolve, highlighted by recent major security incidents targeting Iran’s banking sector.

• Living Security has launched its 2026 Cybersecurity Awareness Month program, providing organizations with a turnkey solution for October security campaigns. • The program offers expert-led learning, interactive experiences, and ready-to-use resources to help companies avoid months of manual content development. • This initiative is specifically designed for highly regulated sectors, such as healthcare and financial services, to reinforce positive security behaviors and navigate trust in an AI-driven environment.

• Anthropic expanded Project Glasswing on Tuesday, adding about 150 organizations in more than 15 countries to its restricted AI security testing program. • The company said its Claude Mythos Preview offensive-security model has surfaced more than 10,000 high- or critical-severity vulnerabilities since April, highlighting rapid gains in AI-assisted defense tools. • Anthropic warned that comparable Mythos-class models could emerge within six to 12 months, a timeline that underscores the short gap between frontier AI capabilities and broader misuse risk.

Announcement by Marco Rubio is being widely seen as setback for Brazil’s president, and a boost for his far-right challengerThe United States has designated Brazil’s two largest criminal gangs, the First Capital Command (PCC) and the Red Command, as foreign terrorist organisations.The announcement, made by Marco Rubio, the secretary of state, on Thursday, is being widely seen in Brazil as a setback for Luiz Inácio Lula da Silva, the president who had strongly opposed the designation – and a boost for Lula’s main challenger in October’s presidential election, the far-right senator Flávio Bolsonaro. Continue reading...