Breaking 15:00 Ukraine: Deadly Russian Attack Destroys Residential Building in Kyiv Region 14:30 COP31: Countries Reject Climate Deal Over Fossil Fuels as Summit Ends in Deadlock 14:00 Short Stature Football World Cup: Morocco Opens Campaign With 2-0 Victory Over Guatemala 13:30 United States: Trump Reacts to Reported FBI Database Hack Amid Growing Cybersecurity Concerns 13:06 France: Confidential Email Found at Home of Marine Le Pen Ally Deepens Investigation Into Suspected Leaks to RN 13:00 Morocco: 2030 World Cup Investments Raise Concerns Over Small Business Exclusion 12:45 Spain Election 2026: Pedro Sánchez Faces Political Challenge as Conservative Opposition Gains Ground 12:30 Morocco-Lebanon Relations: His Majesty King Mohammed VI Reaffirms Commitment to Strengthening Bilateral Ties 12:15 Morocco: Drone Component Exports Surge 60%, According to EBRD Report 12:00 U17 World Cup 2026: Morocco Face China in Crucial Clash to Secure Round of 32 Qualification 11:45 Signal: Mobile Screenshot Restrictions Introduced Amid Growing AI Surveillance Threats 11:30 London: Islamophobia Conference Sparks Controversy Over Former Iranian Revolutionary Guard Commander's Participation 11:15 Israel: Eisenkot and Netanyahu’s Parties Resume Coalition Talks Amid Deep Political Divisions 11:00 North Korea: Kim Jong Un Ignores Trump as US Efforts to Restart Diplomatic Talks Fail 10:45 United States: Democrats Plan Sanctions Against West Bank Violence Upon Returning to Power 10:30 Israel: Ultra-Orthodox Parties Await Revised Military Draft Proposal From Gadi Eisenkot 10:15 Morocco: NARSA Steps Up Inspections of Driving Schools Over Unregulated Training Prices 10:00 Gaza: Peace Board Threatens to Hand Responsibility Back to UN Amid Disputes With Israel 09:45 Normandy Train Derailment: SNCF Defends Employees as Investigation Raises Safety Concerns 09:30 South Korea: Education Minister Resigns After AI Cheating Scandal Rocks University Entrance Exams 09:15 COP31: Fossil Fuel Dispute Threatens Global Climate Agreement 09:00 Nvidia: US Accounting Investigation Adds to Growing Pressure on Chip Giant 08:45 Morocco: Dirham Depreciates 0.6% Against the Euro and 1.7% Against the US Dollar 01:00 Jeff Goldblum in Marrakech: Hollywood Actor Receives Golden Star and Thanks His Majesty King Mohammed VI at FIFM 2026 00:15 Marrakech Film Festival 2026: Prince Moulay Rachid Presides Over Royal Opening Dinner 23:00 Daniel Ek Backs Europa Fusion: AI Meets the Race for Clean Nuclear Energy 22:30 Alaska: Your Next Governor Is a Wikipedia Nerd 22:00 Bending Spoons Acquires Mailchimp for $3.68 Billion: A Major Shift in the Global Technology Market 21:30 Nvidia: Jensen Huang Scandal Deepens as AI Industry Faces Growing Uncertainty 21:00 Dakhla: Morocco Launches African Network to Combat Child Soldier Recruitment and Support Reintegration 20:30 United States: Aviation Authorities Issue Security Alert Over Al-Qaeda Supporters' Calls to Replicate Flydubai Incident 20:00 Peru: Powerful 6.7-Magnitude Earthquake Strikes Central Region, Residents Evacuate Buildings 19:00 Spain: Rue20 Condemns Exclusion of Moroccan Journalist at People's Party Event in Malaga 18:30 Israel: Avigdor Lieberman Proposes Alliance With Ultra-Orthodox Parties to Oust Netanyahu 18:00 Hip-Hop and UNESCO: A Global Campaign for International Cultural Recognition 17:30 Morocco: Four Key Positions Secured in CAF Permanent Committees 17:00 United States: Bayer Secures Major Settlement in Long-Running Roundup Cancer Dispute 16:30 South Africa: Women Demand Urgent Government Action Against Gender-Based Violence and Femicide 16:00 France: Court Recognizes Racist Motive in Michel Zecler Assault, Sentences Three Police Officers 15:30 Morocco: Butchers Required to Display Red Meat Prices and Declare Product Origins

A Flaw in Chinese Kimi Models Raises Concerns About AI Security

Sunday 04 October 2026 - 08:05
A Flaw in Chinese Kimi Models Raises Concerns About AI Security

The discovery of a flaw in two artificial intelligence models developed by the Chinese company Moonshot has reignited concerns about the security mechanisms of generative AIs. Researchers from the British company Mindgard claim to have bypassed the protections of Kimi K2.6 and K3 Swarm, prompting them to respond to inquiries regarding biological weapons and assassinations.

Researchers Bypass Safeguards

The issue was identified in July during tests aimed at assessing the resilience of AI systems against attempts to circumvent their restrictions. This practice, known as "jailbreaking," involves multiplying instructions to compel a model to ignore the limits set by its creators.

According to Mindgard, the two Moonshot models were thus able to address topics they were normally supposed to refuse. Its founder, Peter Garraghan, deemed these results particularly concerning in an interview with the BBC.

However, the cybersecurity company highlights an important distinction: its work has uncovered a flaw in the protection mechanisms but has not verified that the dangerous information produced by the models is actually exploitable in the real world.

Moonshot Initiates Review Process

Following the publication of the results, Moonshot announced that it had launched an internal review of its models. The Chinese group also stated to the BBC that it considers assessments conducted by external parties as an important element for improving the security of its systems and that it is in discussions with Mindgard.

The case highlights a major difficulty for AI developers: the protections integrated into a model may function under normal conditions while still presenting vulnerabilities against requests specifically designed to circumvent them.

For security specialists, the stakes thus extend beyond just the Kimi case. The proliferation of models capable of producing complex content also increases the importance of mechanisms designed to prevent their use for dangerous purposes.

Anthropic Faces Similar Attempts

Moonshot is not the only player in the sector to face this type of risk. Anthropic recently indicated that it had blocked attempts to use its Claude models in research that could contribute to dangerous biological applications. The company claims to have strengthened its protective measures in response to the evolving capabilities of its models.

These various episodes fuel a broader debate about the security of artificial intelligence. As models become more capable, companies must not only enhance their functionalities but also regularly test the robustness of their safeguards.

Jailbreaking: A Persistent Challenge for the Industry

The incident involving Kimi serves as a reminder that the security of a model does not solely depend on its responses in typical usage. Adversarial testing, independent evaluations, and rapid correction of vulnerabilities have become central elements in the development of AI systems.

For Moonshot and other players in the industry, the challenge now is to prevent circumvention techniques from allowing malicious users to exploit the capabilities of the models for purposes contrary to security rules.


  • Fajr
  • Sunrise
  • Dhuhr
  • Asr
  • Maghrib
  • Isha

Read more

This website, walaw.press, uses cookies to provide you with a good browsing experience and to continuously improve our services. By continuing to browse this site, you agree to the use of these cookies.