SAS Institute, a North Carolina-based analytics software provider, has cut 300 positions across the company. The layoffs, implemented last week, affect employees globally across SAS's 400-plus offices worldwide, with its main campus located in Cary. The company stated the cuts align resources with long-term strategic priorities whilst continuing investment in research and development, sales and cloud operations. Affected employees can apply for other open positions or access transition services. SAS currently has several hundred vacancies worldwide. The company emphasised its focus on aligning resources with customer needs and sustainable growth, but declined to speculate on further changes.

Nevada's Division of Social Services has selected SAS to help reduce error rates in its Supplemental Nutrition Assistance Program using AI technology. The state will deploy SAS Payment Integrity for Food Assistance to avoid potential penalties under new federal requirements mandating SNAP error rates below 6%. In the first two months of implementation, caseworkers have already identified $130,000 in monthly overpayments. The solution uses AI and machine learning to analyse every active SNAP case, providing monthly risk scoring to help staff prioritise reviews based on error likelihood and dollar amounts at risk. Nevada faces a potential $52 million penalty if error rates exceed the federal threshold. Beginning in 2027, states exceeding the 6% error rate must pay between 5% and 15% of total SNAP benefits, with some penalties potentially reaching $1 billion annually.

Insurance fraud through AI-generated images is rapidly escalating, with fraudsters using artificial intelligence to create or alter photos depicting crash scenes, receipts and property damage. The threat costs US consumers an estimated $308.6 billion annually, with one in 10 property-casualty insurance losses including fraud. Only 7% of anti-fraud professionals say their organisations are prepared to detect AI-charged fraud, according to a survey by the Association of Certified Fraud Examiners and SAS. To combat this, SAS has developed a synthetic image detection solution combining computer vision, optical character recognition and large language model reasoning. The system provides automated content screening, multi-signal fraud detection and risk-based decisioning. It can identify subtle manipulations that escape human detection and extends beyond insurance to banking and government sectors.