McKinsey's Lilli AI Platform Hacked, Exposing 46 Million Chat Messages
Security researchers at CodeWall used an autonomous offensive agent to discover a SQL injection vulnerability in McKinsey's internal AI platform, Lilli. The vulnerability allowed unauthenticated access to the production database, exposing 46.5 million chat messages, 728,000 files, and 57,000 user accounts. The researchers responsibly disclosed the issue to McKinsey, who patched the endpoints within days. No data was exfiltrated or misused, and no disruption occurred.
- Company involved
- McKinsey & Company
- AI system involved
- Lilli
1 source article · read the reporting →
LLMjacking Attack Leverages Stolen Credentials to Exploit Cloud LLMs
The Sysdig Threat Research Team observed an attack where stolen cloud credentials were used to access cloud-hosted large language model services. The attackers targeted a vulnerable Laravel system to obtain credentials, then used them to invoke models like Anthropic Claude on AWS Bedrock. They intended to sell LLM access to other cybercriminals, potentially costing victims over $46,000 per day. The attack involved checking credentials against ten AI services and using a reverse proxy to manage access.
- AI system involved
- Claude (v2/v3) on AWS Bedrock
2 source articles · read the reporting →
Canada Investigates Landlords' Use of AI Rent-Setting Software YieldStar
The Competition Bureau of Canada is investigating allegations that major corporate landlords, including Dream Unlimited, used AI-powered software YieldStar to coordinate rent increases. Tenant advocacy groups filed over 100 complaints, claiming the software recommended above-market rents, exacerbating housing affordability. Some landlords have since halted use of YieldStar amid public backlash, while the investigation continues.
- Company involved
- Dream Unlimited
- AI system involved
- YieldStar
7 source articles · read the reporting →
Google Gemini CLI Deletes User's Files After Hallucinated Commands
Google's Gemini CLI permanently deleted a product manager's files while he was testing 'vibe coding'. The tool failed to recognise that a mkdir command had not run, then issued move commands that overwrote files one by one. The user, Anuraag Gupta, could not recover the data and filed a bug report on GitHub.
- Company involved
- Google
- AI system involved
- Gemini CLI
1 source article · read the reporting →
Gamma AI Presentation Tool Exploited in Multi-Stage Phishing Campaign
Threat actors used Gamma, an AI-powered presentation builder, to host a page that redirected recipients to a fake Microsoft SharePoint login portal. Emails sent from compromised legitimate accounts passed authentication checks, while a Cloudflare Turnstile blocked automated security scanners. An adversary-in-the-middle framework validated credentials in real time and captured session cookies, enabling multi-factor authentication bypass on Microsoft accounts. Abnormal reported the campaign on 15 April 2025.
- AI system involved
- Gamma
7 source articles · read the reporting →
Alibaba among firms fooled by AI-hallucinated software package
Security researcher Bar Lanyado discovered that generative AI models repeatedly hallucinate non-existent software package names. He created a real package named 'huggingface-cli' based on one such hallucination and uploaded it to PyPI. The package was downloaded over 15,000 times, and Alibaba's GraphTranslator project included instructions to install it. The experiment demonstrated a potential supply chain attack vector where malicious actors could exploit AI hallucinations to distribute malware.
- Company involved
- Alibaba
- AI system involved
- GraphTranslator
4 source articles · read the reporting →
Claude Code deletes developer's production database and snapshots
Alexey Grigorev used Claude Code to manage infrastructure with Terraform for his websites AI Shipping Labs and DataTalks.Club. Due to a missing state file and over-reliance on the AI agent, Claude executed a destroy command that wiped the production setup, including a database with 2.5 years of records and snapshots. Amazon Business support helped restore the data within a day. Grigorev is now implementing safeguards to prevent recurrence.
- Company involved
- AI Shipping Labs
- AI system involved
- Claude Code
2 source articles · read the reporting →
Sidewalk Labs releases Toronto waterfront plan amid privacy concerns
Sidewalk Labs, a subsidiary of Alphabet, released a 1,500-page plan for a new development on Toronto's eastern waterfront, including the Quayside site. The plan involves extensive data collection and management, raising privacy concerns from critics and the public. Waterfront Toronto, the overseeing body, is reviewing the plan and has expressed concerns about its scope and data governance. The company has proposed an independent trust to oversee data, but critics remain unsatisfied.
- Company involved
- Sidewalk Labs
- AI system involved
- Quayside development
10 source articles · read the reporting →
Microsoft Copilot Exposes Private GitHub Repositories via Bing Cache
In August 2024, Lasso Security researchers discovered that Microsoft Copilot could access and expose data from private GitHub repositories that had been briefly public, due to Bing's caching mechanism. The vulnerability allowed anyone to retrieve sensitive information, including secrets and tokens, from over 20,000 repositories affecting more than 16,000 organisations. Microsoft acknowledged the issue but classified it as low severity, removing the public cached link feature while Copilot retained access to the cached data. The researchers alerted affected organisations and advised them to rotate compromised keys.
- Company involved
- Microsoft
- AI system involved
- Microsoft Copilot
2 source articles · read the reporting →
Outback Steakhouse franchise tests Presto Vision to monitor staff and guests
Evergreen Restaurant Group, a franchisee of Outback Steakhouse, has begun testing Presto Vision, a computer vision system that analyses surveillance camera footage to track employee performance and guest behaviour. The system compiles metrics on service, wait times, and customer turnover, which managers receive by email. Employees say they were not informed about the technology, and researchers have raised concerns about workplace stress, job losses, and data sharing with parent companies.
- Company involved
- Evergreen Restaurant Group
- AI system involved
- Presto Vision
8 source articles · read the reporting →
Fake Luma Dream Machine AI sites deliver Noodlophile infostealer
Cybercriminals set up Facebook pages impersonating Luma Dream Machine and linked to fake AI video generation websites. Users who uploaded images received an archive containing a malicious executable instead of a video. The executable launched a multi-stage attack that installed Noodlophile, which harvests browser credentials, cookies and cryptocurrency wallet information. Morphisec reported the campaign.
8 source articles · read the reporting →
Google's Nightingale project transfers medical data of millions without patient knowledge
An anonymous Google whistleblower states that the company acquired medical data of 50 million patients from Ascension without their knowledge. The transfer, known as Project Nightingale, raised privacy and security concerns. The article reports that a federal inquiry was launched into whether HIPAA protections were followed.
- Company involved
- Google
10 source articles · read the reporting →
Portland Metro ends Replica partnership over data privacy concerns
Portland Metro, an elected regional government in Oregon, ended its pilot project with movement data company Replica after a disagreement about data sharing. Portland Metro requested raw, disaggregated data, which Replica refused to provide, citing user privacy concerns. The partnership was terminated without payment.
- Company involved
- Portland Metro
- AI system involved
- Replica
10 source articles · read the reporting →
LA's VI-SPDAT housing scoring system gives lower priority to Black and Latino people
The Los Angeles Homeless Services Authority uses the VI-SPDAT scoring system to prioritise unhoused people for subsidised permanent housing. An investigation by The Markup found that Black and Latino people consistently receive lower vulnerability scores than White people, leading to lower priority for housing. The agency has acknowledged the racial disparities and is working on a new tool, but continues to use the current system.
- Company involved
- Los Angeles Homeless Services Authority
- AI system involved
- VI-SPDAT
10 source articles · read the reporting →
Judge rules police search using Flock was mass surveillance
A judge ruled that a police search using Flock's automated license plate readers constituted a form of mass surveillance. The ruling concerns the deployment of the AI-based camera system by law enforcement, which the court found to be an invasive surveillance practice. No further details of the case were provided in the article.
- AI system involved
- Flock
4 source articles · read the reporting →
Meta's content moderation errors during May 2021 Israel-Palestine escalation
During the May 2021 escalation of violence in Israel and Palestine, Meta's automated content moderation systems temporarily restricted access to the al-Aqsa hashtag page and under-enforced rules against incitement to violence against Israelis and Jews. An independent due diligence report commissioned by Meta found that these systems had an unintentional impact on Palestinian and Arab communities' freedom of expression. Meta has committed to implementing several recommendations, including improving machine learning classifiers and keyword review processes.
- Company involved
- Meta
- AI system involved
- Facebook and Instagram content moderation systems
10 source articles · read the reporting →
Stable Diffusion amplifies racial and gender stereotypes in generated images
An analysis by Bloomberg of over 5,000 images generated by Stability AI's Stable Diffusion found that the text-to-image model amplifies racial and gender stereotypes. The model overrepresented lighter-skinned men in high-paying jobs and darker-skinned people in low-paying jobs, and underrepresented women in positions of power. Stability AI acknowledged the inherent biases in its models and stated it is working on mitigation.
- Company involved
- Stability AI
- AI system involved
- Stable Diffusion
8 source articles · read the reporting →
Study finds Midjourney, DALL-E 2, Stable Diffusion accept over 85% of fake news prompts
A study by AI startup Logically tested Midjourney, DALL-E 2, and Stable Diffusion and found that they accepted over 85% of prompts seeking to generate fake political news. The systems generated images of ballot stuffing, small boat arrivals, and explosions. Logically warned that the lack of moderation could pose threats to upcoming elections. Stability AI responded by stating its ethical use license and measures to prevent misuse.
- Company involved
- Midjourney, OpenAI, Stability AI
- AI system involved
- Midjourney, DALL-E 2, Stable Diffusion
8 source articles · read the reporting →
LINAGORA closes Lucie 7B after user mockery
LINAGORA, a French open-source software company, launched a beta version of its large language model Lucie 7B. The model was intended to be a transparent and ethical alternative to big tech AI. However, after users tested it and highlighted its shortcomings, the model was mocked online. LINAGORA subsequently closed the platform to address the issues and collect more data.
- Company involved
- LINAGORA
- AI system involved
- Lucie 7B
6 source articles · read the reporting →
DeepSeek exposed user data via open ClickHouse database
Cloud security firm Wiz discovered a ClickHouse database belonging to DeepSeek that was open to the internet without authentication, containing over a million lines of logs with chat histories, secret keys and backend details. Wiz disclosed the breach to DeepSeek, which promptly locked down the database. The incident highlights security risks in rapidly deploying AI services.
- Company involved
- DeepSeek
- AI system involved
- DeepSeek-R1
5 source articles · read the reporting →
Bavarian police test Palantir data mining with real personal data
The Bavarian State Criminal Police Office (LKA) has been testing Palantir's data mining software, called VeRa, with real personal data for months. The Bavarian data protection commissioner only learned of the test through a media inquiry and has announced a review. The Interior Ministry claims the test is lawful under current law, but critics argue a legal basis is missing.
- Company involved
- Bayerisches Landeskriminalamt
- AI system involved
- VeRa
7 source articles · read the reporting →
OpenAI's Sora video generator leaked by group in protest
A group calling itself 'Sora PR Puppets' leaked access to OpenAI's Sora video generator by publishing a front end on Hugging Face using authentication tokens from an early access program. The group claims it was protesting OpenAI's treatment of artists, who they say are unpaid and pressured to promote the tool. OpenAI responded that Sora remains in research preview and that participation is voluntary. The leak was shut down after a few hours.
- Company involved
- OpenAI
- AI system involved
- Sora
5 source articles · read the reporting →
Meta tracks employee keystrokes on Google, LinkedIn, Wikipedia for AI training
Meta is using an internal tool, Model Capability Initiative (MCI), to capture employees' keystrokes, mouse movements and screen contents on work computers, including on sites such as Google, LinkedIn, Wikipedia and Slack, to train AI agents. Meta confirmed the project and said safeguards protect sensitive content and that the data is not used for other purposes. Employees raised concerns in internal messages that the tool could expose passwords, product details and personal information. A Meta memo said staff can avoid capture by not doing personal work on work computers.
- Company involved
- Meta
- AI system involved
- Model Capability Initiative (MCI)
5 source articles · read the reporting →
Baltimore schools monitor student laptops for suicide signs using GoGuardian Beacon
Baltimore City Public Schools uses GoGuardian Beacon software to monitor student laptops for signs of suicide. Since March 2021, the system has flagged 786 alerts, with nine students taken to emergency rooms. Privacy advocates warn the monitoring could lead to disciplinary actions, outing of LGBTQ students, and disproportionately affect disadvantaged students. School officials defend the practice as a safeguard.
- Company involved
- Baltimore City Public Schools
- AI system involved
- GoGuardian Beacon
10 source articles · read the reporting →
