Senator Ron Wyden (D-Ore.), with Senator Cory Booker (D-N.J.) and Representative Yvette Clarke (D-N.Y.), introduced in early February the Algorithmic Accountability Act of 2022. This bill aims to bring transparency and oversight of software, algorithms and other automated systems that are used to make automated decisions. "As algorithms and other automated decision systems take on increasingly prominent roles in our lives, we have a responsibility to ensure that they are adequately assessed for biases that may disadvantage minority or marginalized communities," said Sen. Booker. The bill requires companies to conduct impact assessments for bias, effectiveness and other factors, when using automated decision systems to make critical decisions. The bill also gives the Federal Trade Commission (FTC) the authority to require the companies to comply with this bill and to create a public repository of these automated systems.

Following on from the publication of its Feasibility Study in December 2020, the Council of Europe's Ad Hoc Committee on Artificial Intelligence (CAHAI) and its subgroups initiated efforts to formulate and draft its Possible Elements of a Legal Framework on Artificial Intelligence, based on the Council of Europe's standards on human rights, democracy, and the rule of law. This document was ultimately adopted by the CAHAI plenary in December 2021. To support this effort, The Alan Turing Institute undertook a programme of research that explored the governance processes and practical tools needed to operationalise the integration of human right due diligence with the assurance of trustworthy AI innovation practices. The resulting framework was completed and submitted to the Council of Europe in September 2021. It presents an end-to-end approach to the assurance of AI project lifecycles that integrates context-based risk analysis and appropriate stakeholder engagement with comprehensive impact assessment, and transparent risk management, impact mitigation, and innovation assurance practices.

As we gradually emerge from the pandemic with tech more deeply embedded in our daily lives, issues relating to the scope and use of the vast amounts of data being collected, such as data security, and the use of predictive technologies and artificial intelligence (AI) are becoming increasingly important. This rise in prominence has meant that awareness of the'darker' side of AI, as highlighted by mainstream hits on Netflix such as Coded Bias, and specifically the ways in which machine learning algorithms have led to greater discrimination and inequalities is on the rise. Technology itself, and the data we collect on a daily basis, is not inherently'good' or'bad'. However, as humans, we can program our biases into the technology we create. Existing racial, social, and gender biases can be programmed into the algorithms we develop, often unconsciously, or the data that are available to train algorithms, biased in nature, inevitably affects the outputs.

In 1996, philosopher Helen Nissenbaum issued a clarion call concerning the erosion of accountability in society due to the ubiquitous delegation of consequential functions to computerized systems. Using the conceptual framing of moral blame, Nissenbaum described four types of barriers to accountability that computerization presented: 1) "many hands," the problem of attributing moral responsibility for outcomes caused by many moral actors; 2) "bugs," a way software developers might shrug off responsibility by suggesting software errors are unavoidable; 3) "computer as scapegoat," shifting blame to computer systems as if they were moral actors; and 4) "ownership without liability," a free pass to the tech industry to deny responsibility for the software they produce. We revisit these four barriers in relation to the recent ascendance of data-driven algorithmic systems--technology often folded under the heading of machine learning (ML) or artificial intelligence (AI)--to uncover the new challenges for accountability that these systems present. We then look ahead to how one might construct and justify a moral, relational framework for holding responsible parties accountable, and argue that the FAccT community is uniquely well-positioned to develop such a framework to weaken the four barriers.

The problem of human trust in artificial intelligence is one of the most fundamental problems in applied machine learning. Our processes for evaluating AI trustworthiness have substantial ramifications for ML's impact on science, health, and humanity, yet confusion surrounds foundational concepts. What does it mean to trust an AI, and how do humans assess AI trustworthiness? What are the mechanisms for building trustworthy AI? And what is the role of interpretable ML in trust? Here, we draw from statistical learning theory and sociological lenses on human-automation trust to motivate an AI-as-tool framework, which distinguishes human-AI trust from human-AI-human trust. Evaluating an AI's contractual trustworthiness involves predicting future model behavior using behavior certificates (BCs) that aggregate behavioral evidence from diverse sources including empirical out-of-distribution and out-of-task evaluation and theoretical proofs linking model architecture to behavior. We clarify the role of interpretability in trust with a ladder of model access. Interpretability (level 3) is not necessary or even sufficient for trust, while the ability to run a black-box model at-will (level 2) is necessary and sufficient. While interpretability can offer benefits for trust, it can also incur costs. We clarify ways interpretability can contribute to trust, while questioning the perceived centrality of interpretability to trust in popular discourse. How can we empower people with tools to evaluate trust? Instead of trying to understand how a model works, we argue for understanding how a model behaves. Instead of opening up black boxes, we should create more behavior certificates that are more correct, relevant, and understandable. We discuss how to build trusted and trustworthy AI responsibly.

Many boundaries are, and will continue to, shape the future of Artificial Intelligence (AI). We push on these boundaries in order to make progress, but they are both pliable and resilient--always creating new boundaries of what AI can (or should) achieve. Among these are technical boundaries (such as processing capacity), psychological boundaries (such as human trust in AI systems), ethical boundaries (such as with AI weapons), and conceptual boundaries (such as the AI people can imagine). It is within this final category while it can play a fundamental role in all other boundaries} that we find the construct of needs and the limitations that our current concept of need places on the future AI.

Despite achieving incredibly low perplexities on myriad natural language corpora, today's language models still often underperform when used to generate text. This dichotomy has puzzled the language generation community for the last few years. In this work, we posit that the abstraction of natural language as a communication channel (\`a la Shannon, 1948) can provide new insights into the behaviors of probabilistic language generators, e.g., why high-probability texts can be dull or repetitive. Humans use language as a means of communicating information, and do so in an efficient yet error-minimizing manner, choosing each word in a string with this (perhaps subconscious) goal in mind. We propose that generation from probabilistic models should mimic this behavior. Rather than always choosing words from the high-probability region of the distribution--which have a low Shannon information content--we sample from the set of words with an information content close to its expected value, i.e., close to the conditional entropy of our model. This decision criterion can be realized through a simple and efficient implementation, which we call typical sampling. Automatic and human evaluations show that, in comparison to nucleus and top-k sampling, typical sampling offers competitive performance in terms of quality while consistently reducing the number of degenerate repetitions.

We present LaMDA: Language Models for Dialog Applications. LaMDA is a family of Transformer-based neural language models specialized for dialog, which have up to 137B parameters and are pre-trained on 1.56T words of public dialog data and web text. While model scaling alone can improve quality, it shows less improvements on safety and factual grounding. We demonstrate that fine-tuning with annotated data and enabling the model to consult external knowledge sources can lead to significant improvements towards the two key challenges of safety and factual grounding. The first challenge, safety, involves ensuring that the model's responses are consistent with a set of human values, such as preventing harmful suggestions and unfair bias. We quantify safety using a metric based on an illustrative set of human values, and we find that filtering candidate responses using a LaMDA classifier fine-tuned with a small amount of crowdworker-annotated data offers a promising approach to improving model safety. The second challenge, factual grounding, involves enabling the model to consult external knowledge sources, such as an information retrieval system, a language translator, and a calculator. We quantify factuality using a groundedness metric, and we find that our approach enables the model to generate responses grounded in known sources, rather than responses that merely sound plausible. Finally, we explore the use of LaMDA in the domains of education and content recommendations, and analyze their helpfulness and role consistency.

We've seen again and again the harmful, unintended consequences of irresponsibly deployed algorithms: risk assessment tools in the criminal justice system amplifying racial discrimination, false arrests powered by facial recognition, massive environmental costs of server farms, unacknowledged psychological harm from social media interactions, and new, sometimes-insurmountable hurdles in accessing public services. These actual harms are egregious, but what makes the current regime hopeless is that companies are incentivized to remain ignorant (or at least claim they to be) about the harms they expose us to, lest they be found liable. Many of the current ideas for regulating large tech companies won't address this ignorance or the harms it causes. While proposed antitrust laws would reckon with harms emerging from diminished competition in the digital markets, relatively small companies can also have disturbing, far-reaching power to affect our lives. Even if these proposed regulatory tools were to push tech companies away from some harmful practices, researchers, advocates and--critically --communities affected by these practices would still not have sufficient say in all the ways these companies' algorithms shape our lives.

The EU strikes again with a new set of regulations that take aim at the use of artificial intelligence (AI) to address the variety of risks associated with the societal adoption of AI. Like its sibling the General Data Protection Regulation (GDPR), the Artificial Intelligence Act (AIA) actually has teeth, with fines rising to €30 million, or 6% of global revenue. Is the answer to delete all your AI systems to minimize your risk to zero, or continue using AI for a competitive edge? Can you manage the recurring costs required to maintain compliance with the AIA even as the technology itself increases your bottomline? Take the famous UK pub chain JD Wetherspoon, founded by British businessman Tim Martin in 1979 who has been an outspoken critic of the EU and a Brexit campaigner. Their response to personal identifiable information (PII) protection, legislated by the GDPR in 2017, was to delete their entire CRM database.