Govt-backed report raises alarm—frontier AI models can now attack banks with minimal human help

• Proxima Fusion, a leading European stellarator company founded in 2023, has secured new early-stage funding to advance commercial fusion power plants. • A spin-out from the Max Planck Institute for Plasma Physics, the company is utilizing the QI-HTS stellarator concept and breakthroughs from the Wendelstein 7-X programme.

• Intel has announced a planned €5 billion expansion of its semiconductor manufacturing facilities located in Ireland. • This move has increased investor interest in US-listed AI infrastructure multinationals that maintain significant operational footprints across Europe.

• European technology investment in Q2 2026 is experiencing a structural shift as capital organizes around strategic sectors. • Funding models have evolved beyond traditional equity, with a growing reliance on "hybrid capital" including debt, grants, public backing, and strategic capital.

• A Gigamon survey reveals that 66% of financial firms allow AI security automation to operate autonomously, while 77% of these firms have experienced breaches via AI. • The report highlights a growing vulnerability, with 47% of firms reporting an increase in attacks specifically targeting AI and Large Language Model (LLM) deployments.

• The EU and UK have officially attributed a cyberattack on Poland's power grid to Russian spies, specifically targeting GRU leaders Vyacheslav Stafeyev, Ivan Senin, and Ivan Kasyanenko. • The operation allegedly involved collaboration with cybercriminals and a firm called IMPULS to recruit cybersecurity specialists from Russian universities to orchestrate hybrid attacks.

• The UK and European Union have implemented their first joint cyber sanctions package, targeting 24 Russian individuals and entities. • These sanctions address Russia's "reckless" efforts to sow division across Europe through malicious cyberattacks and the spread of deceptive anti-Ukraine narratives.

• India’s second Digital Threat Report for the BFSI sector reveals that six out of seven emerging cyber threats predicted last year have already become fully operational. • Attackers are leveraging artificial intelligence, stolen identities, and manipulated payment workflows to create fraudulent activities that closely mimic legitimate transactions.

• Singapore is rapidly deploying AI agents across government and business sectors, but this acceleration is introducing significant security vulnerabilities. • Risks such as misconfigured agents, flawed prompts, and manipulated instructions are identified as primary threats that could disrupt critical business operations.

• Microsoft CEO Satya Nadella has initiated a strategic shift by warning companies to protect themselves from dominant AI giants like OpenAI and Anthropic. • This move signals a pivot in the relationship between big tech infrastructure providers and the AI labs they previously championed.

• Senior banker and investor relations professional Gulamaly Hussain cautioned that AI should support, rather than replace, investor judgment during a virtual interaction with startups, SMEs, and students. • Hussain acknowledged that while AI has significantly increased the speed and accessibility of investment research, it lacks the essential human discipline and sound financial thinking required for complex decisions.

• AI leaders from Anthropic and OpenAI are calling for the creation of a global oversight committee to monitor advanced AI systems and forcibly slow development if recursive self-improvement poses a risk. • Recent algorithmic breakthroughs in detecting vulnerabilities in high-security cybersecurity software have triggered widespread anxiety across both public and private sectors.