Meta Platforms Inc., the parent company of Facebook and Instagram, has announced changes to its content policies regarding AI-generated content. Under the new policy, Meta will no longer remove misleading AI-generated content but will instead label it to provide transparency. This shift in approach aims to address concerns about misleading content without outright removal.
Previously, Meta’s policy targeted ‘manipulated media’ that could mislead viewers into thinking someone in a video said something they did not. Now, the content policy extends to digitally altered images, videos, or audio as the company will employ fact-checking and labelling to inform users about the nature of the content they encounter on its platforms.
The policy was revised in February after Meta’s Oversight Board criticised the previous approach as ‘incoherent’. The board recommended using labels instead of removal for AI-generated content, and Meta has agreed with this perspective, emphasising the importance of transparency and additional context in handling such content.
Why does it matter?
Starting in May, AI-generated Meta-platform content will be labelled ‘Made with AI’ to indicate its origin. This policy change is particularly significant given the upcoming US elections, with Meta acknowledging the need for clear labelling of AI-generated posts, including those created using competitors’ technology.
Meta’s shift in content moderation policy reflects a broader trend toward transparency in dealing with AI-generated content across social media platforms. By implementing labelling instead of removal, Meta aims to provide users with more information about the nature of the online content.
According to an investigation by Israeli publications +972 Magazine and Local Call, the Israeli military has reportedly utilised AI technology, known as Lavender, to assist in selecting bombing targets in Gaza. The AI system was allegedly developed following attacks by Hamas on 7 October. Lavender was said to have identified around 37,000 Palestinians in Gaza as suspected ‘Hamas militants’ for potential assassination.
Based on the interviews with Israeli intelligence officers, the Lavender system was used without thorough independent examinations of the identified targets before airstrikes. The system analysed data associated with known Hamas operatives to rank other individuals in Gaza on a 1–100 scale for similarity. However, concerns were raised about the loose definition of ‘Hamas operative’ used during the system’s training, leading to errors in target identification.
The AI-driven warfare strategy reportedly allowed significant collateral damage, with wide latitude given to intelligence officers regarding civilian casualties. During the conflict, officers were purportedly authorised to cause civilian deaths as a part of targeting suspected Hamas operatives. Additionally, a system called ‘Where’s Daddy?’ was allegedly used to track targets to their homes, resulting in bombings that sometimes killed entire families, even when the targeted individual was not present.
Why does it matter?
The use of AI technologies like Lavender and facial recognition programs in Gaza has raised ethical concerns, especially as they contribute to civilian casualties. Mona Shtaya, a non-resident fellow, highlighted these technologies’ extension of Israel’s regional surveillance efforts. Reports of civilian deaths due to AI-targeted strikes underscore the need for careful consideration of the ethical implications of using advanced technologies in conflict zones.
OpenAI has unveiled new features to enhance model performance and customizability, catering to developers seeking to optimise AI implementations for speed, accuracy, and cost-efficiency. These enhancements include a self-serve fine-tuning API for GPT-3.5, which has already facilitated the training of hundreds of thousands of models across various organisations since its launch in August 2023. Fine-tuning empowers models to grasp content deeply and augment their existing capabilities, resulting in superior outcomes tailored to tasks such as code generation, text summarisation, or personalised content creation.
One standout success story involves a leading job-matching platform that utilised fine-tuning GPT-3.5 Turbo to refine personalised job recommendations for users. By leveraging fine-tuning, it achieved an 80% reduction in token usage, allowing the platform to scale its messaging volume from under a million to roughly 20 million monthly messages, enhancing user engagement and satisfaction.
In addition to the fine-tuning API, OpenAI has introduced new features to provide developers with enhanced control and visibility over their fine-tuning projects. These features include epoch-based checkpoint creation, a comparative playground for model evaluation, third-party integrations with platforms like Weights and Biases, comprehensive validation metrics, and improved dashboard functionality.
Why does it matter?
The ChatGPT’s owner envisions widespread adoption of customised models across industries and use cases. It emphasises the importance of effectively scoping use cases, implementing robust evaluation systems, and leveraging the right techniques to optimise model performance.
OpenAI accommodates organisations seeking fully custom-trained models capable of comprehending complex domain-specific knowledge and behaviours. The company has expanded its Custom Models program, offering assisted fine-tuning services for organisations that require tailored AI solutions beyond the capabilities of standard fine-tuning. Assisted fine-tuning involves collaborative efforts with technical teams to implement advanced techniques, such as parameter-efficient fine-tuning, to maximise model performance across specific use cases or tasks. Thus, for organisations seeking to harness AI capabilities for personalised impact, OpenAI’s offerings provide a pathway towards tailored and effective AI implementations.
A bipartisan proposal in the US aims to bolster border control by integrating cutting-edge technologies such as AI, machine learning, biometrics, and nanotechnology. Spearheaded by the Department of Homeland Security (DHS), the legislation mandates developing a comprehensive plan within 180 days to incorporate these technologies into border security operations. The move follows the release of an AI Roadmap for DHS and an executive order emphasising trustworthy AI for American benefit.
Representative Lou Correa highlighted the importance of investing in security-enhancing technologies to aid Customs and Border Protection (CBP) officers in swiftly responding to threats like human trafficking and hazardous migrant crossings. The proposed plan includes metrics, performance indicators, and privacy/security assessments to ensure effective implementation.
As cartels and foreign adversaries increase in sophistication amid the ongoing border crisis, the necessity to deploy advanced technologies becomes apparent. The legislation seeks to leverage commercially available technologies while empowering CBP Innovation Teams to adapt and integrate them into border security operations efficiently.
The bill also mandates that the CBP clarify operational procedures and roles regarding new technologies. Research areas outlined in the legislation encompass mobile surveillance vehicles, lighter-than-air ground surveillance equipment, tunnel detection, and other pertinent areas determined by the Secretary of Homeland Security. Through bipartisan efforts, the proposal aims to equip officers with the tools necessary to safeguard the border effectively.
Artifact had previously decided to wind down its app operations, citing limited market opportunities. However, Yahoo sees potential in Artifact’s AI-powered recommendation engine and other features to bolster its news operations. This acquisition aligns with Yahoo’s existing portfolio, which includes news brands like TechCrunch, Engadget, and Yahoo Finance, as well as a minority stake in Taboola, a content recommendation platform.
Yahoo, now under the ownership of private equity firm Apollo Global Management since 2021, aims to leverage Artifact’s technology to deliver personalised content and scale its news offerings. The co-founders of Artifact, Kevin Systrom and Mike Krieger, who previously worked at Meta (formerly Facebook), will advise Yahoo during the transition. Their departure from Meta in 2018 was reportedly due to differences in vision with Meta CEO Mark Zuckerberg.
Yahoo did not disclose the financial details of the acquisition. However, the incorporation of Artifact’s technology signals Yahoo’s commitment to innovation in the competitive landscape of digital media.
OpenAI unveiled a new AI tool, Voice Engine, capable of generating lifelike speech by analysing a mere 15-second audio sample, as announced in OpenAI’s blog post. The tool aims to offer reading assistance, aid translation efforts, and provide a voice for nonverbal individuals with speech conditions. Despite its potential benefits, OpenAI acknowledges the serious risks associated with the technology, especially during an election year.
Voice Engine, developed by OpenAI in late 2022, has undergone private testing with a select group of partners who have agreed to usage policies requiring explicit consent from original speakers and prohibiting unauthorised impersonation. OpenAI stresses the importance of transparency, with partners mandated to disclose that the voices are AI-generated and that all audio produced by Voice Engine includes watermarking for traceability.
OpenAI advocates for responsible deployment of synthetic voices, suggesting measures such as voice authentication and a ‘no-go voice list’ to prevent misuse. The company recommends phasing out voice-based authentication to access sensitive information like bank accounts. However, the widespread release of Voice Engine remains uncertain as OpenAI seeks feedback and evaluates the results of its small-scale tests before deciding on broader deployment.
Why does it matter?
The introduction of Voice Engine comes amid rising concerns over AI-generated deepfakes and their potential to disseminate misinformation, particularly in political contexts. Recent incidents, such as a fake robocall imitating President Biden and an AI-generated video of a Senate candidate, underscore the urgency of addressing advanced AI technologies’ ethical and societal implications.
As concerns grow over the proliferation of AI-generated disinformation, schools and lawmakers are doubling down on media literacy education. The push, already underway in 18 states, aims to equip students with the skills to discern fake news, which is particularly crucial as the 2024 presidential election looms. Beyond politics, the harmful effects of social media on children, including cyberbullying and online radicalisation, underscore the urgency of these efforts.
States like Delaware and New Jersey have set the bar high, mandating comprehensive media literacy standards for K-12 classrooms. These standards promote digital citizenship and empower students to navigate media safely. Yet, disparities exist among states, with some, like Illinois, implementing more muted forms of media literacy education, focusing primarily on high school instruction.
In response to the lack of federal guidelines, bipartisan efforts in Congress, such as the AI Literacy Act, seek to address the gap. Introduced by Rep. Lisa Blunt-Rochester and Rep. Larry Bucshon, the bill aims to integrate AI literacy into existing education programs, emphasising its importance for national competitiveness. However, progress on the bill has stalled since its introduction, leaving the federal approach to media literacy uncertain.
Despite variations in implementation, students across states are embracing media literacy education positively. For educators like Lisa Manganello in New Jersey, the focus is on fostering critical thinking and information literacy, irrespective of political affiliations. As misinformation continues to increase online, the need for media literacy education at the state and federal levels remains paramount to empower students as responsible digital citizens.
In the 350th session taking place in March, the agenda includes a High-level Section on ‘the challenges and opportunities of digitalization, including artificial intelligence (AI) and algorithmic management, for the world of work’. Academic, industry, and intergovernmental body experts are invited to shed light on digitalization’s impacts on labour, e-government in support of labour and social protection, and the role of ILO in enhancing the benefits of digitalization.
For more information, please refer to the official page.
A New York-based startup, Hume AI, unveiled a groundbreaking AI voice interface, the Empathic Voice Interface (EVI), designed to mimic human emotions and engage in more natural conversations with users. This emotionally intelligent AI can be integrated into various applications, spanning customer service, healthcare, and virtual and augmented reality environments. Following a successful funding round of $50 million led by EQT Ventures, the beta version of EVI has been launched to the public.
Unlike conventional AI systems guided by superficial human ratings, Hume AI’s approach involves training the AI to understand user preferences and generate vocal responses optimised for user satisfaction. By learning directly from proxies of human happiness and continuously updating its knowledge with each interaction, EVI aims to reconstruct human preferences from first principles, thus enhancing its ability to engage with users effectively.
Its empathic Large Language Model (eLLM) is central to Hume AI’s innovation, which integrates multimodal generative AI with expression measures. This model enables EVI to adjust its language and tone based on context and the user’s emotional expressions, creating a more personalised and engaging conversational experience. The voice model was trained on extensive data from millions of human interactions, enhancing its ability to effectively understand and respond to users’ needs.
Why does it matter?
While Hume AI leads the charge in emotionally intelligent AI, other researchers and institutions also explore similar avenues. Columbia University’s Hod Lipson has developed an AI-powered robot capable of predicting human expressions and replicating them. At the same time, scientists from South Korea’s Ulsan National Institute of Science and Technology have created a facial mask that records verbal and non-verbal expression data. These innovations hint at a future where technology seamlessly integrates human emotions into various applications, from virtual reality concierges to personalised services based on users’ emotional states.
OpenAI, the Microsoft-backed company in AI, is gearing up to establish its presence in Tokyo this April, marking its foray into Asia as it expands its global operations. This move comes after the successful establishment of offices in London and Dublin last year, according to a source familiar with the matter who preferred to remain anonymous due to the sensitive nature of the information.
The decision to set up in Japan underscores the growing significance of the Asian market in AI development and adoption. OpenAI’s presence in Tokyo signifies a strategic move to tap into Japan’s burgeoning interest in AI technology, with major players like SoftBank Corp. and Nippon Telegraph and Telephone Corp. actively venturing into AI-driven services tailored for Japanese speakers.
The AI giant’s expansion plans in Japan have been in the pipeline for some time, with discussions intensifying after a meeting between OpenAI’s CEO, Sam Altman, and Prime Minister Fumio Kishida last April. Altman expressed the organisation’s intent to bolster its Japanese language services and collaborate with the government to address potential risks and establish regulatory frameworks.
Why does it matter?
OpenAI’s decision to establish a foothold in Tokyo reflects its commitment to catering to the Japanese market’s evolving needs and its broader strategy to engage with international stakeholders and foster AI innovation on a global scale. As the demand for AI-powered solutions continues to surge worldwide, OpenAI’s move into Tokyo signals a significant milestone in its quest to shape the future of AI.
