Flock Safety

Flock Safety's AI-powered gunshot detection technology implemented in San Jose has been reported to yield a high rate of false positives, incorrectly flagging sounds such as fireworks or cars backfiring as gunfire. Initially, only 50 percent of the detected incidents were confirmed as gunshots. After recalibration, accuracy improved, suggesting that such systems may not be as reliable as claimed. The system's potential to dispatch police to non-threatening situations raises concerns, especially in communities of color, about the risks of unnecessary police confrontations.

Source: https://www.wired.com/story/ai-gunshot-detection-accuracy-san-jose-nyc/

TPRM report: https://scoringcyber.rankiteo.com/company/flock-safety

"id": "flo450070624",
"linkid": "flock-safety",
"type": "Vulnerability",
"date": "6/2024",
"severity": "50",
"impact": "3",
"explanation": "Attack with significant impact with internal employee data leaks"

{'affected_entities': [{'industry': 'Technology',
                        'location': 'San Jose',
                        'name': 'Flock Safety',
                        'type': 'Company'}],
 'description': "Flock Safety's AI-powered gunshot detection technology "
                'implemented in San Jose has been reported to yield a high '
                'rate of false positives, incorrectly flagging sounds such as '
                'fireworks or cars backfiring as gunfire. Initially, only 50 '
                'percent of the detected incidents were confirmed as gunshots. '
                'After recalibration, accuracy improved, suggesting that such '
                "systems may not be as reliable as claimed. The system's "
                'potential to dispatch police to non-threatening situations '
                'raises concerns, especially in communities of color, about '
                'the risks of unnecessary police confrontations.',
 'impact': {'brand_reputation_impact': ['Concerns about Reliability',
                                        'Potential Risks to Communities of '
                                        'Color'],
            'operational_impact': 'High Rate of False Positives',
            'systems_affected': 'Gunshot Detection System'},
 'lessons_learned': 'AI systems may not be as reliable as claimed and can lead '
                    'to unintended consequences.',
 'post_incident_analysis': {'corrective_actions': 'Recalibration of Algorithm',
                            'root_causes': 'Inefficiency in AI Algorithm'},
 'recommendations': 'Continuous monitoring and recalibration of AI algorithms '
                    'to improve accuracy and reliability.',
 'response': {'remediation_measures': 'Algorithm Recalibration'},
 'title': "Flock Safety's Gunshot Detection System False Positives",
 'type': 'System Malfunction',
 'vulnerability_exploited': 'AI Algorithm Inefficiency'}