OpenAI's Security Incident and Japan's $16 Billion AI Chip Investment
OpenAI confronts a security setback, SiFive secures a $2 billion valuation boost, and Japan invests $16 billion in AI chip innovation.
EDITION #136, Apr 12
OpenAI confronts a security setback, SiFive secures a $2 billion valuation boost, and Japan invests $16 billion in AI chip innovation.
Image Source: Reuters
Key Developments in AI & Security this Week
This week, significant cybersecurity developments emerged in the AI sector. OpenAI reported a security issue linked to a third-party tool, Axios, which affected its macOS app signing workflow. Although no user data was compromised, OpenAI is requiring all macOS users to update their apps by May 8 to mitigate risks from counterfeit applications.
Meanwhile, Goldman Sachs is enhancing its cybersecurity measures in response to concerns regarding Anthropic's AI model, Mythos. Following a meeting with U.S. regulators, CEO David Solomon emphasized the need to bolster defenses against evolving cyber threats posed by advanced AI systems.
- OpenAI's Axios-related security issue revealed no user data was accessed, but updates are mandatory by May 8.
- Goldman Sachs is upgrading its cyber defenses after warnings from regulators about Anthropic's Mythos AI model.
- Anthropic's Mythos could exploit vulnerabilities in legacy systems, raising alarms in the banking sector.
- The FBI shut down a phishing operation, W3LL, that targeted over 17,000 victims worldwide.
Source: The Verge
SECURITY ALERT
OpenAI Faces Security Incident
OpenAI has identified a security issue linked to a compromised version of the Axios tool used in its macOS app-signing process. On March 31, 2026, a malicious Axios version was introduced, potentially affecting apps like ChatGPT Desktop and Codex. Fortunately, OpenAI confirmed that there is no evidence of user data being accessed or altered.
In response, OpenAI is updating its security certificates and has mandated that users update to the latest macOS app versions. Older versions will cease receiving updates after May 8, 2026, to ensure software integrity and user safety.
Image credit: TechCrunch
AI CHIP VALUATION
SiFive's Valuation Soars to $3.65 Billion
SiFive, the open-chip-design company behind RISC-V, has successfully raised $400 million in a funding round led by Atreides Management, boosting its valuation to $3.65 billion. This funding, which saw participation from Nvidia among other notable investors, marks a significant shift for SiFive as it expands from embedded chips into AI data-center CPUs.
With this capital, SiFive aims to enhance its offerings by integrating its open designs with Nvidia’s CUDA and NVLink Fusion, positioning itself to compete in the growing AI infrastructure market.
Image source: Bloomberg
JAPAN'S AI CHIP PUSH
Japan's $16 Billion Bet on AI Chips
Japan is significantly ramping up its AI chip ambitions with a $16 billion investment aimed at supporting Rapidus, a startup focused on advanced semiconductor manufacturing. The government recently approved an additional ¥631.5 billion ($4 billion) in subsidies, bringing total public funding to approximately ¥2.6 trillion ($16.3 billion) by March 2027. This funding is crucial for Rapidus as it develops cutting-edge 2-nanometer logic chips, with Fujitsu as its initial client.
This strategic move underscores Tokyo's commitment to enhancing Japan's position in the global semiconductor market, despite the inherent risks. An external committee has already recognized Rapidus' progress at its Hokkaido foundry, reflecting confidence in the project's potential to reshape Japan's role in AI technology.
Source: Reuters
REGULATORS ON ALERT
UK Regulators Scrutinize Anthropic's AI Model
UK regulators are rapidly assessing risks linked to Anthropic’s new AI model, Mythos. The Bank of England and Financial Conduct Authority are collaborating with the National Cyber Security Centre to evaluate vulnerabilities identified by Mythos, which reportedly detected thousands of security flaws across various systems. This urgent review will involve major banks like Goldman Sachs and JPMorgan Chase, focusing on the implications for financial stability and cybersecurity.
As part of this initiative, regulators are pushing for stronger defenses against potential AI-driven cyber threats. The scrutiny comes amid concerns that Mythos could enhance the capabilities of cybercriminals, raising alarms about the security of interconnected financial networks.
PAYMENT CUTS
X Reduces Payments to Clickbait Accounts
X is slashing payments to accounts that rely heavily on clickbait and rapid news aggregation. This includes a 60% cut in payouts for aggregators this cycle, with an additional 20% reduction planned. Nikita Bier, X's head of product, emphasized that this move aims to curb manipulation of the monetization program and foster genuine creator growth.
The decision follows complaints from conservative news accounts about demonetization, raising concerns about the impact on consistent content creators. X's strategy highlights a shift towards prioritizing quality over quantity in content creation.
