Risk Channel delivers the latest, most relevant and useful business intelligence to key decision makers and influencers, each weekday morning.

Nvidia and dozens of other tech companies including Microsoft, SpaceX and Palantir have launched a new artificial intelligence safety initiative focused on open models. The announcement follows news that a Chinese open-weight AI model was able to defend against a cyberattack carried out by rogue OpenAI models. The target of the attack, startup Hugging Face, was unable to use leading U.S. frontier models to defend itself, because guardrails did not distinguish between aggressor and defender. Instead, it turned to a self-hosted, open-weight Chinese model, which was not bound by those same rules. “The Open Secure AI Alliance will work to remediate and disclose vulnerabilities using open technologies,” Nvidia said. “The recent Hugging Face security incident delivered a clear reminder: cyber defenders need open, frontier agentic systems for self-defense.”

 

CNBC

Strengthen Cyber Defences Against Emerging AI Threats

AI is reshaping the threat landscape, creating new risks that traditional detection models may struggle to identify.

This report from Lenovo explores the leading AI-driven threats facing IT and risk leaders, and outlines how organisations can strengthen their cybersecurity approach. Learn why existing defences must evolve, where vulnerabilities may emerge and how a more proactive strategy can help protect the modern workplace.

Designed for technology, cybersecurity and risk management professionals, the report provides practical insight into preparing for a rapidly changing security environment.

Download the report

 

Frontier AI models are dramatically accelerating the discovery of cybersecurity vulnerabilities, creating new challenges for banks whose remediation processes and governance frameworks were not designed to operate at machine speed. As the volume of identified vulnerabilities grows, financial institutions face increasing pressure to fix critical exposures before they can be exploited. Experts encourage banks to focus on four priorities to strengthen cyber resilience: prioritizing vulnerabilities based on exploitability and business impact rather than severity alone, scaling remediation through AI-enabled testing and cross-functional collaboration, reducing risk while vulnerabilities remain unpatched through measures such as zero-trust architecture and layered defenses, and modernizing governance to enable faster, real-time decision-making during cyber incidents.

 

Wall Street Journal

Microsoft has announced its first artificial intelligence model designed to identify vulnerabilities in source code. When integrated with OpenAI's GPT-5.4, MAI-Cyber-1-Flash reportedly outperforms competitors including Anthropic's Mythos 5 and Google's 3.5 Flash Cyber. “We have world-leading performance at 50% of the cost,” said Mustafa Suleyman, CEO of Microsoft AI. The initiative marks a significant push in Microsoft's cybersecurity strategy, led by Hayete Gallot, who in February rejoined the company as executive vice president of security.

 

Techcrunch

CNBC

More than half of corporate boards have not established rules governing the use of artificial intelligence, according to a Deloitte survey, highlighting potential risks related to legal liability, data security, and the handling of confidential information. The survey found that 51% of boards lack AI policies, 47% do not use AI in board activities, 25% permit AI use without standardized tools, and only 8% rely on company-approved AI platforms for committee work. The findings suggest that board-level adoption of AI remains in its early stages, with governance practices still evolving. Where policies do exist, they typically focus on security, confidentiality, acceptable use, legal compliance, and recordkeeping. Deloitte also reported that 77% of boards have held AI education or briefing sessions for directors within the past six months, while only 10% have taken no steps to improve AI knowledge.

 

Yahoo Finance

Johnson & Johnson has agreed to pay up to $5.5bn to settle outstanding lawsuits alleging its talc-based products caused ovarian cancer, marking a significant step towards ending more than a decade of litigation. The settlement is conditional on 95% of remaining claimants opting in and would resolve nearly 80,000 claims. The company expects to pay up to $3bn in 2027, with no further payments due before 2028. The agreement follows recent court rulings that weakened the plaintiffs' cases, after expert witnesses were unable to directly attribute individual cancers to talc. J&J has consistently denied the allegations and said the claims lacked scientific merit.

 

CBS News

Financial Times

The Globe and Mail

Wall Street Journal

Cracker Barrel has announced that chief executive Julie Masino will step down next month after less than a year in the role, following a failed rebranding effort that sparked customer backlash and failed to reverse weakening sales. David Deno, the former chief executive of Bloomin' Brands, will take over on August 10th, while Masino will remain in an advisory role until October 9th. The restaurant chain's attempted modernization, including a new logo and redesigned stores, was quickly reversed after customer criticism, but trading has remained weak, with same-store sales declining 1.8% in its fiscal third quarter ended May 1st.

 

CNBC

New York Times

US News and World Report

The Washington Post

Uber has cut 10% of jobs within its customer service operations team, saying it is working “to simplify operations, strengthen in-person collaboration, and continue to embrace AI.” Employees who had been working remotely for the team were also asked to relocate to a hub office, consistent with the company’s return-to-office mandate, an Uber spokesperson said. “Our organization has become too complex and siloed,” Megha Yethadka, Uber’s vice president of global community operations, wrote in a memo. She said the department “has made some strides” in using AI, “but to unlock this potential, we need an effective organization to layer AI on. We cannot scale frontier technology on top of fragmented processes.”

 

Bloomberg

Engadget

Israeli technology giant monday.com has announced a restructuring that will result in the laying-off of 620 employees, or about 20% of the company's global workforce. The decision is part of a product transformation aimed at developing an AI-based work platform. The company said it is flattening its organizational structure to enhance decision-making speed and efficiency. “This is the most painful decision we have made since founding monday.com – yet we are certain it is the right one. We made it. We own it. And we take full responsibility for it,” Monday.com co-founders Roy Mann and Eran Zinman wrote in a letter to employees. “The organization we built for our previous chapter is not the organization that fits the new AI era.”

 

The Jerusalem Post

The Times of Israel

The U.S. ​Food and Drug Administration's Center for ​Veterinary Medicine is launching ​a pilot program to encourage domestic manufacturing ‌of animal health drugs and reduce the risk of supply chain disruptions. The new pilot program is intended to reduce the risk of such disruptions by prioritizing reviews of animal drug applications where both finished drug products and the active ingredients are manufactured in the U.S.

 

Feedstuffs

Reuters

Blackstone says the pace of redemptions at its flagship private credit fund had slowed, after a surge in customers worried about potential losses seeking to withdraw their money this year.

 

Financial Times

Flávio Bolsonaro, a son of former President Jair Bolsonaro, has secured his party's nomination to challenge incumbent President Luiz Inácio Lula da Silva in the upcoming October elections. Despite endorsements from international figures including Argentina's Javier Milei and Israel's Benjamin Netanyahu, Flávio's campaign has been marred by scandals and a lack of domestic support. He is still searching for a running mate and faces challenges in forming alliances. His campaign has been further complicated by self-inflicted problems, including a recent admission that he had requested millions of dollars from jailed banker Daniel Vorcaro to pay for a film about his father's life. 

 

Los Angeles Times

The New Jersey Supreme Court has established new ethics rules barring lawyers from engaging in sexual relationships with clients they represent. Sexual relationships predating a legal representation are permissible, but lawyers cannot begin new intimate relationships with a client they represent, the court said. Attorneys representing a company or other entity are also prohibited from creating new sexual relationships with individuals in the company who are part of the organization's “litigation control group” that oversees legal disputes on which the lawyer works.

 

Bloomberg Law

Law.com