The Efficient & Responsible AI R&D Summit, hosted at Intel’s Haifa campus, brought together senior executives from leading technology firms and researchers from Israeli and international academia to address the challenges posed by the next generation of artificial intelligence.
The conference focused on the rapid development of large language models, autonomous agents, and advanced computing architectures, with particular attention to risk management as AI systems increasingly make real-world decisions.
Speakers emphasized that efficiency gains must be balanced with responsibility, transparency, and clear regulatory frameworks. “This can only be achieved through collaboration between academia and industry,” said Prof. Eduard Yakubov, president of the Holon Institute of Technology (HIT), warning against the accelerated adoption of AI technologies without adequate oversight.
Technology leaders, AI researchers, and data science experts presented case studies and practical implementations, highlighting innovations in responsible AI while fostering dialogue between sectors. The summit was led by Intel, HIT, and IAHLT, and initiated by Dr. Nava Shaked, head of HIT’s School of Multidisciplinary Studies, together with Avner Algom, founder of the AI Academia-Industry community.
Among the guests of honor were Dr. Sarah Basson, head of Accessibility, Disability and Inclusion at Google Global, and Prof. Constantinos Petrides, vice rector for internationalization at the Mediterranean University of Cyprus. Basson stressed that artificial intelligence does not eliminate human bias. “Humans build the algorithms, collect the data and decide what to do with the results,” she said, noting that visual AI systems can still exclude entire populations, often without organizations realizing it.
She added that accessibility is a core element of Google’s corporate responsibility. Elad Shtrit, senior vice president for R&D at Rafael, described a shift toward AI-driven autonomous systems in defense technology. “We’re moving from a world where an operator manually flies a drone to one where agents operate entire fleets,” he said, adding that the key question remains how much trust can be placed in independent decision-making systems.
Infrastructure challenges were also discussed. Tomer Sasson, vice president at Intel’s Personal Computing Division, emphasized the growing importance of edge computing over reliance on the cloud, citing reliability, control, and energy efficiency as critical factors.
Closing discussions highlighted growing awareness that irresponsible AI deployment can lead to legal, reputational, and societal harm. Participants agreed that academia plays a central role in developing ethical standards, assessing bias, and training future AI professionals, while industry cooperation is essential to ensure that AI systems are not only powerful but also fair and trustworthy.